Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
Nishith Rastogi (Locus) shares the engineer's guide to SaaS scale up
0 Plays2 seconds ago
From a $10,000 first customer to multi-million dollar contracts, hear the incredible growth story of Locus!
In this episode of the Founder Thesis podcast, Founder and CEO Nishith Rastogi shares key strategies on how a shift from B2C to B2B helped fuel its success in the on-demand economy. Gain key insights into their go-to-market strategy, AI integration, and the importance of building strong customer relationships. Get actionable insights for your own entrepreneurial journey!
Akshay Datt, a serial entrepreneur who has run ventures in employability training and hiring, hosts the Founder Thesis podcast. He has interviewed 500+ founders to date.
Connect:
Nishith Rastogi: https://www.linkedin.com/in/nishithrastogi/
Akshay Datt: https://www.linkedin.com/in/akshay-datt/
Founder Thesis: www.founderthesis.com
Recommended
Transcript
Nurture vs. Nature in Success
00:00:00
Speaker
America was found with a short heard around the world. Locus was found when there was no answer on the stack. So the idea was to, you know, get together with amazing people and build almost impossible If you see the movie King Richards, you will see their father had a plan of making them a tennis superstar before they were born. There's a psychiatrist, Hungarian psychiatrist by the name Mr. Polga, and he effectively committed to making his offsprings a chess grandmaster before they were born.
00:00:28
Speaker
They ended up earning the highest ever ELO ratings for a women chess player. Nurture, upbringing, your circumstances, the five people around you who have a very meaningful effect in life.
00:00:40
Speaker
You know, the problem with interviewing so many founders is that you eventually realize how any formula and any advice that someone gives you, there is an example of somebody who did not follow the advice and yet succeeded.
00:00:55
Speaker
A very common advice you would receive is that for a startup to get funded by VCs, you must have a tech co-founder and a sales and marketing co-founder, and that makes a good team. But today I'm speaking to Nishit, the founder of Locus, and he is one of two co co-founders, both of whom are hardcore techies, engineers, geeks who like to solve problems using code and technology.
00:01:15
Speaker
And yet they have built a business with multimillion dollar ARR, global revenues, average contract values of a million dollars a year. So which means one single customer is paying them a million dollars a year for their software.
Diverse Paths to Startup Success
00:01:28
Speaker
their way of building the business, how they did it was extremely first principles and counterintuitive, which is why I love this conversation.
00:01:36
Speaker
I'm your host, Akshay Dutt, and you're listening to the Founder Thesis Podcast. Welcome to the Founder Thesis Podcast. We were just chatting a couple of minutes back when you said the
00:01:51
Speaker
to the founder thesis podcast ah we were just chatting um couple of minutes back when you said the in the nurture versus nature debate, you come on the side of nurture. ah What do you mean by that?
00:02:06
Speaker
so So, you know, during parenting, there's ah there's a huge discussion. And I recently became a parent, you know, I had my first child six months back, was blessed with her.
00:02:17
Speaker
right and So I again started discussing it with my father who is a psychiatrist. And there's something we've been debating for a long time, whether you know ah whether in a child it's nature or whether it's nurture.
00:02:28
Speaker
Right. And I think also as a founder mentality, someone who likes to believe that my circumstances are a manifestation of my own actions. As a family, we are very certainly in the nurture camp.
00:02:41
Speaker
And then there are good examples about it. So very famously, for example, if you see the movie King Richards, right, based on the Serena and Venus Williams life, you will see their father had a plan of making them a tennis superstar before they were born.
00:02:57
Speaker
thanks Similarly, there was there's a psychiatrist, a Hungarian psychiatrist by the name of Mr. Polga. He's a big proponent of this debate, and he went into on the nature side nurture side.
00:03:09
Speaker
And he effectively committed to making his offsprings a chess grandmaster before they were born. The Polgar sisters not only went on to become the chess grandmasters, they went on to become the current world champions and not just current world champions, they ended up earning the highest ever ELO ratings for a women chess player, historically.
00:03:31
Speaker
right So yeah, we strongly believe that you know nurture, upbringing, your circumstances, the five people around you have a very meaningful effect in your life. What made you an entrepreneur?
00:03:43
Speaker
what kind of nurturing made you an entrepreneur?
00:03:48
Speaker
ah My parents are doctors and they were very encouraging. I was i was a kid. I was the little kid with a screwdriver in his hand, right? I would like literally go around the home disassembling things, inspecting them, right?
00:04:03
Speaker
And they were extremely supportive of that. right my My father was a psychiatrist in a government hospital, so we didn't come from a lot of money, right? And yet I never heard anything about, you know, oh, if you have damaged this, oh, you have damaged that. So they were extremely supportive of me exploring things, me trying out new things.
00:04:23
Speaker
They were extremely supportive of me not getting going into medical science and being more into the engineering side. i So for me, entrepreneurship is primarily a proxy to building things.
00:04:35
Speaker
right And i my parents definitely encouraged me to be a builder. right the they They got me the right tools. you know ah we had a We had a computer before we had an air conditioning in our house.
00:04:48
Speaker
Wow. and i and And I was raised in Delhi. And I think that's a tough sacrifice for a family to make. Right. And I think that went that went on a long way in establishing what are priorities for me.
The Evolution of Locus
00:05:01
Speaker
Right. Encouraging me being a builder. And that's how I actually like and ended up becoming a hardware hacker. And in Locus office, we still have a hardware lab, even though we're a software enterprise software company.
00:05:12
Speaker
When did you a hardware hacker? like this This was in Bitspilani. Your engineering was from Bitspilani. My engineering was in Bitspilani, but this I'm referring to more around my high school days, right? So during my high school days, I used to build bots interfaced with a parallel port because that was the most simple port. Then we moved on to USBs and all.
00:05:30
Speaker
Yeah, so this is i'm more talking around my high school days and pre-engineering. So and from bits... Yeah. No, I was just saying that, you know, I went to a school in Delhi, Mothers International, which was right opposite the IIT Delhi campus.
00:05:46
Speaker
And IIT Delhi used to have their tech fest with robotics events and all. And that was like my first taste of it. And then I got addicted to it. Fascinating. ah So from bits, you joined Amazon. ah I am assuming Amazon would have also had some role in your in shaping you as an entrepreneur because that culture is known to be famously entrepreneurial.
00:06:10
Speaker
Yeah. So Amazon had a significant role in ah not whether we become an entrepreneur or not, but once we became an entrepreneur, what were some of the things we also incorporated in our organization, right?
00:06:26
Speaker
So Amazon has 14 principles, you know, which they nurture in every joining. Some of them we have definitely ported to Locus as well. So for example, one of them is bias for action, right?
00:06:39
Speaker
That don't don't suffer from a bystander effect, even if it's outside your role, you know, take action. right Similarly, another one is disagree and commit, where effectively, you know you can have a lot of disagreement during the stages of discussion.
00:06:56
Speaker
But once we have all combined reached a rational discussion, then everyone's committed towards making that decision you know success. Nobody can say, oh, but initially I said something else. right So Amazon had a very, yeah, exactly. No, I told you so. Right. Very well said. Right. So just, ah I think a more professional articulation of that.
00:07:16
Speaker
And that has become a very, very meaningful tenet at Locust, right? Because in a startup, you're often making with sufficient data, but still incomplete data decisions.
00:07:28
Speaker
Okay. ah What made you quit Amazon and, you know, start something on your own?
00:07:38
Speaker
So often people think that, you know, it's because I was missing something in my job, right? It was actually quite the opposite, right? I was very fortunate to both work in the AWS team. Then within the AWS team, I was working on the AWS machine learning team.
00:07:53
Speaker
Then within the and AWS machine learning team, I was working on the science aspects of it, right? Almost all of my colleagues were PhDs. Half of my colleagues were SCDs or had a second PhD, right? What is ah it? SCD?
00:08:09
Speaker
doctorate of science so this is like after phd okay okay right or they had like a second phd right uh so it was a it was a brilliant academic environment there were lots of great discussions right and uh we also kind of had like a card branch on the budget so it was it was it was a great life the uh i was more on the research side there were less delivery pressures and all on so there was a lot of learning opportunities right uh What was the use of this machine learning? Was it for powering the e-commerce arm of Amazon? or like
00:08:46
Speaker
So both. So we were working on both building the algorithms for the AWS machine learning platform itself, as well as this group used to run a small, effectively like an internal machine learning consulting for any other group in Amazon who would want to.
00:09:01
Speaker
So for example, one of the problems we were working on that time was standardizing shoe and slipper sizes across brands.
00:09:10
Speaker
to have better customer experience, so things like that. Then, of course, catching fraud and so on and so forth. But we were also working on the platform team. But at that point, ice you know so one of the motors at Locus is to invent the future with amazing people.
00:09:26
Speaker
thank And ah at that point, I thought that, you know, we could still work on more future forward stuff. We could still work on things which everyone, you know, wants, but says it's impossible currently.
00:09:37
Speaker
And effectively, that's what, you know, made me quit Amazon to start my own. Wanting to start my own had also been a longstanding
Route Optimization and Logistics
00:09:46
Speaker
desire. And that's why when we, the first hiring post of Locus was ah America was found when, with a short heard around the world, Locus was found when there was no answer on the stack overflow.
00:10:00
Speaker
So the idea was to you know get together with amazing people and build almost impossible things. So just to translate this for a non-techie, no answer on stack overflow means... a that code has not been written for that particular back is word scaling platform. Code sharing thing, right?
00:10:20
Speaker
Yeah. In the pre-GPT world, right? The chat GPT have slashed as traffic by half, but yeah, the pre-GPT world, yes. Yeah. Now, instead of going to Stack Overflow, you'll just ask chat GPT to write the code for whatever you're looking.
00:10:34
Speaker
it's a bit different, right? ah i'm i'm um ah I'm an absolute lover of what these LLMs and chat GPTs have done to, you know, for programmers, right?
00:10:47
Speaker
Because it's actually made programming a lot of fun. you can you you know ah So right now, if previously a great engineer was a 10x engineer, now a great engineer is clearly a 100x engineer. ba what What it has made hard is to be an average engineer.
00:11:08
Speaker
right That's where it's competing. But if you were a good engineer, of a great engineer, right you could define things, you could delegate out. Now you now we see even within the organization, you know one person finishing things which would have typically taken three good folks about a week.
00:11:23
Speaker
And a lot more time is spent now in understanding the customer requirements and a lot of boilerlor boilerplate stuff is templatized out. So yeah. Fascinating.
00:11:34
Speaker
So, okay. ah What was it? You know, I would like to quote Ben Affleck here, right? And he articulated it really well that ai is a craftsman and it's very important to have good craftsmen, right?
00:11:49
Speaker
A craftsman knows how to do what, right? And AI as a craftsman can typically do more than a human do, right? But the art in the human is to know when to stop.
00:12:05
Speaker
So that's how you have to look at the whole LLM piece that it's a great craftsman and you are the art that brings meaning into the project. Okay. So you need to define and define it really well, what you want to do and how you want to do.
00:12:18
Speaker
And now instead of, you know, struggling by about hiring craftsmen and all, you now have like a mechanized craftsman who can help you execute things. Right. Okay. Okay.
00:12:29
Speaker
ah So, you know, coming back to the birth of Lucas, what was that journey like? So you quit Amazon. Did you have a product idea in mind that I want to build this? What was version one that you built? How did you sell it?
00:12:47
Speaker
Yeah, ah the first version of Locus had nothing to do with Locus. It was a it was ah product called RideSafe and it had completely ah risen out of my personal needs.
00:13:00
Speaker
So let's dial back to 2014, you know, ah further, I would say around the end of 2014, um We had a infamous incident in India, the Nirbhaya incident, and a couple of them similarly. And these incidents were close to us.
00:13:20
Speaker
ah My father was actually in involved as a post-traumatic specialist in some of them. I also have an elder sister. So this topic just got very close to the bone. thank And at that time, there used to be this advertisement about an app where you know if you shake the app, it will send an SOS alert.
00:13:39
Speaker
and There was a very famous Bollywood star who would advertise it for a product. And I thought that app was a security theater. What I mean by that is it creates a false sense of security. right Because if you're in a cab and you are in trouble and you know you are in trouble and you can actually get to your phone, why would you shake it? Why would you not press redial or something?
00:14:02
Speaker
The problem that happens is you typically don't know you are in trouble by the time you're already in a trouble. in you know So what we did was we analyzed patterns, right? with the With the help of my father and a few public officials, we got on past data, we analyzed patterns and we realized in all of these incidents, the one common thing is that the driver takes the victim to an unsafe location.
00:14:23
Speaker
That location is typically not on your route. like So we went on to build this small algorithm called real-time route deviation detection. It abbreviates to R2D2, so it was a tribute to that.
00:14:35
Speaker
What it would do is say you're leaving from airport and you put your home destination. right Now, regardless of what exact way you take, you know, sometimes you want to take your favorite way, one favorite way, two. The app constantly monitors your current location and takes an intelligent decision if it is in the direction to your home or, you know, towards an unsafe place.
00:14:54
Speaker
And then it alerts you. If you don't react, it makes your phone ring. And that's how, you know, then the driver knows that, you know So we have actually had few, you know, but few true positives on it.
00:15:05
Speaker
And that always helps me you know sleep really well whenever I'm troubled. So that's how we started Locus. But during 2015 start, if you remember, everything was becoming click a button and get me something. right So you had on-demand coffee, on-demand flowers, on-demand food. right Both Bangalore and San Francisco were like you know assisted living communities for youth. You press a button and something happens.
00:15:30
Speaker
For that something to happen, it needs complex logistics behind But these were built by product experts, not logistics experts. So the same algorithms that we used for tracking the location, monitoring the location, determining whether it's safe, what is the right route, we transposed all of them into an enterprise package.
00:15:49
Speaker
And we became a one-stop shop for these startups who wanted to create an on-demand service to connect from their order management and make sure their customers are getting a great experience. So that's how through the 2014 and 15 we pivoted into an enterprise app starting from a very personal B2C application.
00:16:08
Speaker
Was ah ah monetization a reason to do the pivot? but I'm sure r two d two would not have been monetizable. Absolutely.
00:16:18
Speaker
And I would like to actually at at first, you know thank my investors toward it. They put us no pressure towards monetization, right? if and You raised one for R2D2. Pardon? but ah You raised funds for R2D2?
00:16:33
Speaker
Yes. ah So after no after we had launched RightSafe, Manish Singhal, who now runs Pi Ventures, he saw me present RightSafe in one of the startup forums and he spotted us.
00:16:46
Speaker
And that's how we went into our fundraise journey. As we raised these funds, our investors were very clear that, you know, this is a great product, but we don't know what's the exact way it will be monetized.
00:16:58
Speaker
But right now, here is some money for both of you to keep on building. So by that time, it was just my co-founder and I, we had hired no team members. like And as we found this opportunity that you know we can actually transform this B2C app into this B2B application for this entire on-demand economy, they were extremely
Global Expansion and Client Engagement
00:17:17
Speaker
supportive. And at this point, Sheetal from GrowX, and now he runs Mirac Ventures.
00:17:22
Speaker
He joined it and we converted the B2C idea into B2B during that fundraise itself. And that's how we did our first 250K back in 2015. Who was the first customer you sold it to?
00:17:37
Speaker
Well, the first customers we sold it to was Urban Ladder and Fresh Menu. Okay. Okay. and The furniture startup Urban Ladder. Yeah. Yeah. Both of whom were running their own delivery services. Yes, back then. Yes.
00:17:51
Speaker
So we we we were the one optimizing it for them. ah This was a product or a service that you sold? this This has always been a product. okay and In our first version, if anything, it was just a business to develop a product where we were only primarily selling APIs.
00:18:09
Speaker
Then we realized our clients actually prefer an end application. And then we became like a proper enterprise SaaS product. So we always provide primarily only a product. And we also don't get into operations.
00:18:22
Speaker
That's our client's business. So for example, we would work with somebody like BlueDart, but we will not start our own transportation company. The business to developer API that you were selling, what was it doing?
00:18:35
Speaker
What was the input and output? Yeah. So the business to developer APIs were that you know if you want to build your own routing systems, own tracking systems, you could use us as an infrastructure play.
00:18:46
Speaker
So during that time, there was Twilio you know who was doing this for telecom authorities. There was Stripe who was doing this for banking authorities. right But we realized when we sell an API stack, we are effectively only abstracting like, you know, the Android location manager or the iOS location service.
00:19:03
Speaker
And then, end and you know, we are not really abstracting out a tough regulatory authority for you to talk to, which is what, say, Stripe or Twilio was doing with banking and telcos, right?
00:19:15
Speaker
We were not doing that. So it made more sense for us to convert into an end application over a pure infrastructure play, right? And that's where, so today we are effectively, you know, one of the leading transport management systems specializing in last mile.
00:19:31
Speaker
Okay. but What, just help me understand, you said infrastructure for managing logistics was what the API was providing. What what does that mean? Like what are what are the components in that?
00:19:43
Speaker
i So let's say if you wanted to do routing, right, we would provide you an API where you can upload all the locations and we will give you a route. But there was no screen to visualize it. There's no screen to edit it.
00:19:56
Speaker
right Then there are no screens to configure. Output would be numbers, right? Input are numbers, outputs are number in a sequence of the things. But the user of that thing then further needs to take those numbers, build a dashboard, build a mobile app, give it to the end user of the driver, right? Because that output is not end consumable by the user of the product.
00:20:17
Speaker
It may be consumable by the buyer of the product, right? So in our space, we you know in our experience, it made sense to make more an application. Then once we were making an application, we realized given we are an optimization system over just a system of records, we save in percentages.
00:20:37
Speaker
If you are saving in percentages, it makes sense to have a large absolute base to save on, right? So that the absolute value is sufficient enough for you to indulge in change management.
00:20:49
Speaker
And hence, then when we started becoming an application, we decided to become a purely enterprise application, having the right security standards, having the right global deployments. Okay.
00:21:00
Speaker
ah what What are the other components? I want to kind of go into the nuts and bolts. So one is and root optimization. ah What else, what are the other components? You said real-time tracking was something you provided.
00:21:14
Speaker
Definitely. Right. So we started with route optimization and that was our Trojan horse, right? That was our core kernel. That was our, you know, competitive edge. We had, we had and have the best on-ground route optimization in the industry, it's widely recognized as such.
00:21:32
Speaker
What makes it best? Three things. One, how closely it can model your real world. So even before speed or anything, right? Because the problem in route optimization is not that, you know, I have, let's have a hundred trucks and thousand packages, what to do, right?
00:21:50
Speaker
Let's take a very simple example of distributing wedding cards. So i want I want to come to Mr. Nipun who lives next to Mr. Akshay. But Mr. Nipun is available only at nine am But Akshay will be available to me only after 4pm post his poker game.
00:22:05
Speaker
Now what should I do? Should I ping Akshay to reschedule? Should I send my aunt to deliver at 9am and we come at 4am? Or should I make a route that starts at 9am there ends back at 4pm here?
00:22:20
Speaker
right Oh, but when I've decided to do that, I have realized that there's also Mr. Agarwal that I should give sweets to. But Mr. Agarwal lives in an area where only a two-wheeler, it's actually convenient to go.
00:22:33
Speaker
like But from there, I need to also go to Nishit. And Nishit lives in an area where you know if I take a non-electric vehicle, I have to pay a surcharge. right And these are only illustrative examples.
00:22:45
Speaker
Your actual production use cases have often over 100 constraints, right? And it actually sounds like a rap song. So the biggest problem in a route optimization engine is to actually model the real world scenario.
00:22:58
Speaker
And then the second and third factors, how fast you can get your output and then how optimized is your output, right? Obviously, so those are the three things that makes. And that was our starting point.
00:23:09
Speaker
Then once we made the route optimization engine, we needed to know whether it's working. How do you know it's working? You have to track its execution. So we built an execution engine. We built an execution engine, we added tracking to it.
00:23:22
Speaker
But then we realized what's what what what's our mission? Our mission is to move any person or any package between two points without human decisions. right So why should you have a person sitting on top of a control tower?
00:23:35
Speaker
So we built a great alert infrastructure. you know So even we have this you know god view mode where sometimes you know we'll kick our legs, put it on the table, have a beer, and just see like bunch of thousands of vehicles moving from here and there.
00:23:49
Speaker
but is that really useful? It's cool, but is that really useful? What is someone looking at looking at that screen? Can we automate that itself? So then we built one of the best control tower interfaces where you know you can not only monitor everything or you can pre-code your actions, you can connect them with your next systems.
00:24:08
Speaker
like As we did this, we realized there's a lot of stuff that happens between an order when when when you you know purchase an order it's ready to be dispatched to its actual dispatch, right? There is quality control that happens. There is documentation that happens.
00:24:26
Speaker
Maybe multiple orders get combined. Maybe the order gets weighted, right? So there's a lot of stuff that happens within the four walls of a hub between the point your order has been decided to be dispatched and it's actually dispatched.
00:24:40
Speaker
So then we build the entire managing systems for that. Then some of our leading customers like Unilever, where we operate in more than nine countries, they started asking us that, hey, you do this last mile piece really well.
00:24:52
Speaker
Can you start jumping into n minus one? So we started building the mid-mile and the first-mile interfaces. As we did this, one of the core architectural changes we had to do was we had previously designed the system for any one leg of the journey.
00:25:08
Speaker
like So now we became a true multi-leg system. right Effectively, we are now a full-fledged competitive transport management system, starting from the point your order is created.
00:25:22
Speaker
And now, actually, for some of our retailers, we are going even before that. Have you ever noticed that when you're checking out on Amazon, sometimes whether you check out or not check out is dependent on the delivery options, right? If you're getting it tomorrow, you may check out, right?
00:25:36
Speaker
So your checkout today is delivery link. So we have now started building delivery link checkout and promise engines, right? So effectively on your product page itself, we will power in low latencies that by when can you get it?
00:25:49
Speaker
What are the various channels you can get it, right? Because today both websites are like real world stores as well as your stores are like your real life website, right?
00:26:00
Speaker
A customer journey in retail could start offline, finish online, could start online, finish offline. And it all needs to be managed across a seamless logistics.
00:26:11
Speaker
So that's effectively the value that we now have morphed into, right? So we operate with clients for whom last mile is important, but we can take a large part of their supply chain from the distribution side.
00:26:23
Speaker
We don't do anything in the procurement side. Okay, fascinating. um I'm going to go a little deeper. Yeah, so this gives you the breadth and now we can you know start zooming into the each areas.
00:26:34
Speaker
Yeah, yeah, yeah, yeah. I love going into the weeds. I should warn you.
Adapting to Industry Changes
00:26:40
Speaker
Amazing. so As a person responsible for product engineering, that's music to my ears.
00:26:45
Speaker
Okay, good to hear that. um The base, ah the Trojan horse is route optimization for which what you do better than anyone else is real world constraints.
00:27:01
Speaker
Modeling. Modeling, yes. Real world modeling. How did you get the data? How did you get the data that this is a two wheeler zone, for example, or this is a zone where if you're not going in an EV, then you have to pay a surcharge or whatever. like How did you get that data?
00:27:16
Speaker
Absolutely. right And there's no silver bullet. So we will scan map. We have multiple maps licenses. We spend seven-digit dollars on them annually.
00:27:30
Speaker
We latch onto municipal data. We also, at any given day, you know probably during the course of this conversation, would have visited another quarter million unique locations and probably more than a couple million locations just today.
00:27:47
Speaker
effectively giving us, you know, over a billion new locations every year. When you say we would have visited, it means that... But but it's the software. Yeah, exactly. The local systems.
00:27:59
Speaker
Right, right, right. Yeah.
00:28:03
Speaker
Right. So there's a lot of execution data that you derive it from. And then the third, fourth and the most important source of data, your clients. Your clients know what they want.
00:28:14
Speaker
They didn't have a system before to model it. Let me give you a great example in Mumbai. Right. So in Mumbai, you have this Andhiri line, right, the the railway line running through the city.
00:28:27
Speaker
Right. And pre what previously the systems would do is that, let's say if you're crossing that line, they will add seven minutes of transaction time so that at least you're waiting. So that was step one.
00:28:39
Speaker
right But if you have a route where the driver is constantly going left, right, left, right to it, they get frustrated. Like, why am I delivering only six items if the other is delivering 16?
00:28:50
Speaker
Even if they're getting paid the same. right So there's attrition. so we actually had to come up with a system that you can in your route you will only cross the line once because if you're delivering to this side of the home it's fine if you're delivering it you will go once deliver all of them and come back we will never create a route which is like this this this right and sometimes we would actually have you know even printouts of local maps and we would ask our clients to mark right so for example in mexico city the core central arterial road every Saturday and Sunday for two hours becomes a cycling track.
00:29:26
Speaker
So you've got to divert all the shipments away from that road, right? Otherwise you will never have accurate ETS, right? Similarly, now we've also built systems where we scan for local news, local weather, sporting events. right So for example, if you know there's a big match that has ended outside Chhatrabaja Shivaji Stadium, we would mark it as a high traffic area.
00:29:49
Speaker
It would do the same if Red Sox finished something at the Fenway Park. OK, fascinating. ah What does this... You know, ah so Locus sits on effectively adding another 9 in everything. What I mean by that is, it roughly takes the same effort to do anything to the first 90%. Then from 90 to 95.
00:30:10
Speaker
Then from 95 99. Then 99 to it becomes useful in production. then from ninety nine to ninety nine point nine yeah but at three nights it becomes useful in production Because you know if you're dispatching 40,000 orders you know every three hours from our dispatch center in Southeast Asia, 0.1% of orders, that's like 400 orders. What do I do about them?
00:30:31
Speaker
right That's 400 angry Twitter messages to sort for. Yeah, yeah, yeah. Fascinating. Okay. ah The execution engine is like a big part of your data collection as well then from what I understand. So is it like absolutely how if you are getting a route through Google Maps, it allows you to report if there is a problem in the route.
00:30:51
Speaker
yeah So like we we don't, yeah, we don't ask for manual interventions. we derive it automatically using accelerometer and GPS data. Because you know we have commercial drivers using our product.
00:31:02
Speaker
They are extremely tight on bandwidth. Our entire product is designed to reduce the number of interactions they have to do over increase.
Client Retention Strategies
00:31:10
Speaker
So the mobile app will show them the route, which is powered by you? The mobile app shows them the route, which is found by us.
00:31:17
Speaker
Okay, they don't need to use Google Maps as such. And sometimes you would also offload it off to the Google Maps, right? Our value is not, you know, telling the from point A to point B to go.
00:31:30
Speaker
Our value is telling what is the point A and what is the point B. And if you're going from point A to point B, should you be avoiding something specific based on which you would have calculated the original? Okay.
00:31:41
Speaker
Okay. Okay. And then if there is a, like, if he's at one place for five minutes, then automatically your system wouldn't pick up that there is something. And then sometimes, you know, we can generate an alert. Sometimes we can let it be, you know, sometimes we can count it as a call. So even even things like, you know I'll give you another example, which our drivers really like that in every other system, often their lunch break is configured, say between one to two or whatever, some fixed time window.
00:32:07
Speaker
But they're like in practice, that's not how I do it. I have my lunch anytime between 12 to 3 when my last job finishes and my next job has some time. So now that's how we route So you can give us a time window in which to slot an hour lunch.
00:32:22
Speaker
And then what we'll do is in that window, whenever a job is finishing, we'll wait, we'll add that slack and then add to the time to the next and then create the optimized routes. So drivers can feel the satisfaction of having, right? Otherwise what happens is they're like, I'm having lunch and I know that order is getting delayed, right? It's just a stressful situation.
00:32:42
Speaker
right Because not that they were not, right? Because we because the product never empathized with the user. It only empathized with the buyer, right?
00:32:53
Speaker
But your adoption will never work if you don't empathize with the user. How did you empathize with the drivers? that You have no context of a driver's life. we um ah Even before we wrote the first line of code, I would have...
00:33:07
Speaker
Yeah, i would have trailed hundreds, if not thousands of and all drivers. I, you know, in every city, there are these chai sutta corners where all the drivers assemble, you know.
00:33:20
Speaker
I spent countless hours there, you know. if On these long roads, you will often see, you know, drivers sometimes sleeping under the truck because that's the only shaded kind of thing. Yeah, I've been a part of that entire thing.
00:33:32
Speaker
Oh, wow. Yeah. And then once we did, once we obviously launched the product, then I would have easily done 1000 plus deliveries for all my clients. Right. So, yeah, it's a, it's a, it's a lot of direct user. yeah as ah If anything, like, you know, i keep telling my, especially my UX team that your primary purpose is to be the voice of the user in the system. Right. Like that guy has no voice.
00:33:56
Speaker
we We get a lot of very, very human feedback sometimes, Akshay. Like, you know, I and remember in this one now one warehouse once in Bombay, right right next to the airport, I went and typically I take off my jacket when I'm entering a warehouse, but that time I was still wearing my dinner jacket and all. So, and I was looking at the Locust product.
00:34:17
Speaker
So, somebody understood that probably I'm someone senior from Locust. And this guy tentatively approached me and I'm like, hi, can I, right? And he basically said that Because now, you know, his sorting time is reduced and the routes are, he is actually able to finish his route on the determined on on the specific time.
00:34:35
Speaker
He's able to take his 7pm local over the 9pm and he finally gets home to see his kid awake. So after four years of his kid attending school, he knows which class he is in.
00:34:46
Speaker
What subjects he's studying. Yeah, that was that was amazing to hear, right? Because those people don't have a voice in the system, right? It's very easy to not have transparency over them for them to not get their fair pay.
00:34:59
Speaker
You see the driver as your primary user?
00:35:05
Speaker
No, he's one of the key users, but we have three sets of users equally important. One is our dispatcher, right? So dispatcher plus customer service. So for us- Dispatcher is the guy in the warehouse? that guy Yeah, accountable for dispatching the routes. And these days, what we have seen in many of our clients is those dispatchers have actually been elevated into you know customer service and customer relationship folks.
00:35:27
Speaker
And then the third is the management for whom we have the entire network design, the analytics, the leaderboards, the cost reports. The control power. Yeah. yeah yeah ah Do you also help them in... Not just the control tower, right? That's on the just just monitoring the execution, but also the decision making.
00:35:49
Speaker
Think of it this way. Today, you know, say which driver often gets the profit most profitable route that traditionally goes between the relationship of the warehouse dispatcher and the driver.
00:36:02
Speaker
That is not good for anyone. Not the organization, not the rest of the drivers, not the transporter, except for those two individuals. Right? Similarly, then within the organization itself across eight dispatch centers, I may have, you know, a different methodologies.
00:36:19
Speaker
I may not have a central I may not even want to have central but I may want to have a central control over each of those eight methodologies. So today with your analytics team and our ah routing team, you can sit and create those routing configurations and options which will then work not just throughout the countries but in larger accounts throughout multiple countries some places you may be doing your own delivery some places it's outsourced right.
00:36:41
Speaker
So it's a it's a great tool to have centralized control and visibility over massively distributed operations. ah yeah In fact, you know, my question was going to be somewhat related to this.
00:36:55
Speaker
There are many different ways in which drivers can be compensated. You could have ah vendor who has a single truck, like a driver come owner, or you could have a driver who's paid per parcel, or you could have a driver on a fixed salary plus incentive, or you could have a vendor who has multiple vehicles to whom you are offloading.
00:37:14
Speaker
ah How do you help companies deal with this complexity?
00:37:20
Speaker
We offer them an option to run all. We have learned the heart, you know, uh, and you take the data costing, how much is each mode costing?
00:37:34
Speaker
Yeah. Everything should be made as simple as possible. That's only a half quarter. The entire court is everything should be made as simple as possible, but no simpler. Okay.
00:37:47
Speaker
So on our first day, we tell all our clients that we are a non-opiniated platform. Why? Because now we work with large enough clients that just one client across their multiple states or multiple countries are operating in completely different models of logistics.
00:38:05
Speaker
So at one place where they have captive logistics, where they're running their own trucks, their drivers may also be on their payroll. right with an overtime charge. So our thing would be to make sure that we are bringing back the drivers within the time.
00:38:21
Speaker
Or to utilize all the trucks because you know the drivers are anyway on the payroll. So minimize the operational hours or minimize the number of trucks. That same country in another thing could be doing it at a per package level where they don't have enough density.
00:38:37
Speaker
right Then they in the third state may be operating where they lease a truck,
Enterprise Sales Insights
00:38:42
Speaker
right? And then decide what should be the route of the truck. And then you are being paid by the distance and thing.
00:38:48
Speaker
So our platform allows you to manage each of these modes and many more. So at the time of onboarding, you kind of collect all this data, like the different constructions. It's not a day zero product, right? It's an enterprise product with an account management.
00:39:04
Speaker
There's a lot of integrations as well, right? with With some of our large clients, our integrations often take over a couple of months, right? And we would integrate into SAPs. Yeah, the we we have integrated into stuff written in IBM mainframe, COBOL, flat files, ELDs.
00:39:23
Speaker
Okay, fascinating. So ah do you also... ah Help them manage vendor invoicing? Yes, we have just, now we have you know, i should i should I should hire you as my product manager.
00:39:37
Speaker
First of all, prior to this, no one has asked, you know, when you say that you love getting getting into weeds, a lot of people say this, but you actually demonstrated that by going into how drivers are incentivized, right? Like that's the that's a genuine attempt at understanding the problem statement.
00:39:54
Speaker
So I first called you out there. um Second, now you took it to ah absolutely right point of finance, corner so we limit ourselves on not dispersing the actual payment.
00:40:07
Speaker
right So we are not a payment system. But now because you know, effectively, which vendor should you use that should also be dependent on your cost contract. Because let's say a vendor may seem expensive for today, but you have not met its monthly minimum quota.
00:40:24
Speaker
So if you don't use that vendor instead of spending 10 extra rupees today, you might end up spending 200 rupees at the month end on the penalty quota. So we actually factor that also now into our cost calculations on how should you be routing and how should you be routing involves with who should you be routing.
00:40:41
Speaker
But now I know what's exactly in your head. That's the LHS, the left hand side of the equation. So then once the execution is done, right, and let's say we predicted a root X, but the driver had to take X dash for whatever reasons, right?
00:40:54
Speaker
And their billing may be a bit different billing. So they will upload an invoice. We will also then help you reconcile these invoices. eight And there we are now adding more and more intelligence that you know previously it had to be a very structured invoice.
00:41:08
Speaker
Now we are all going over to scanning paper invoices. And yeah, that's the idea that because what we learned from working with now tier one public market companies, even in India, is that nobody wants to really delay vendor payments.
00:41:22
Speaker
That's a you know that's our mentality of the past. Most of our clients are growth companies. Their problem is that they don't have efficient, you know reliable transportation vendors.
00:41:34
Speaker
If to their good vendors, they pay on time, they stick with them, even over a discount. right So many of my tier one customers in India, all of them public market companies, have explicitly told me that Nishit, regardless of whatever you hear in the market, our primary aim is to pay our transportation vendor on time. If I can pay them on N minus one day, I want that. I don't want to make a float on this amount.
00:41:58
Speaker
What is costly for me is not to have a transportation vendor and that impact my sales. Interesting. So logistics seems to be supply start market. Absolutely. okay Maybe eventually the vendor need not even do invoicing outside of Locus. That's the idea. That's the idea, right? You finish it, it generates, everyone accepts it, it's done.
00:42:21
Speaker
Okay, okay, okay. Fascinating. Okay. ah So, okay. ah Help me understand what is the control tower? What does that do, like, you know from the management perspective?
00:42:33
Speaker
Maybe not from management perspective, right? So control tower is what helps you in your transactional stuff, right? So fundamentally control tower is monitoring about life, right?
00:42:45
Speaker
So let's say if I'm operating in three countries, let's let's take a three countries, let's take a small footprint, say 20 states, 10,000 vehicles, right? ten thousand vehicleres but ah One of my dispatch centers haven't dispatched out anything one hour after they had to be dispatched.
00:43:07
Speaker
How would I know? Because typically, you know, yeah you you understand, right? Like from someone who has to know from the someone who has to take an action, there are probably eight people involved in this chain.
00:43:19
Speaker
Assuming each of them have an immediate intent to also inform, it will still take over one day, right? Or till the customer escalates. So first, control tower helps you ah get alerted.
00:43:32
Speaker
That was phase one. But now, if you have been using our control tower for a few months or a few years, we also allow you to encode your actions. That let's say for a B2C company, if my customer order is getting delayed more than five minutes, but less than 15 minutes, send them an apology message with a 5% discount code.
00:43:53
Speaker
So that can also be automated back. right So the control tower is primarily now involved into delivering exceptional customer experience. Right. So for example, even even the interface right now, one of the things I really hate when I call any customer service center is they ask me nine different things, right? Your name, your this ID or that ID. I'm like, you you know, the number I'm calling you from, right? Like make some intelligent guess.
00:44:20
Speaker
Right. So exactly how you have in Chrome, where you type on the ah URL bar, any part of the yeah URL or search term it brings up, that's what we now have on our control tower. That you can literally type anything about that order and all the matching so that you can quickly get to the order and inform the customer. Because see, the customer has already skipped past the tracking link, right? So the customer is already troubled.
00:44:41
Speaker
That time, don't try and make make a ration card for the customer. That's right. Okay. Okay. Okay. And got it. Okay. You also power the data which the customer gets, like a tracking link, which they get that information on that tracking link. Like say on Amazon, it shows you to track your order. so So stuff like that also you would power for your brand. So we get a tracking link. That tracking link actually doubles up as a few things, right? It also works up as a safety mechanism in certain countries for certain products.
00:45:12
Speaker
That tracking link has the face and name of the driver detail, right? So the customers are used to only opening if it's matching. Right. That customer link also has an ETA and then it can also have an OTP. So if you want like a handshake delivery, that customer link, yeah, could ah you could also sign and then that becomes, you know, the moment you sign a package, you will get an email with your signature embedded and proof of delivery.
00:45:37
Speaker
Right. Then ah proof I'll give you a very cool example of you know how today AI is getting embellished in itself. right So often you would see drivers taking a photo, leaving a package outside the gate right as a proof of delivery.
00:45:54
Speaker
Now in that, they're well they' are always trained to take the number of the house in that thing. So now what have started doing is automatically started extracting out that number.
00:46:05
Speaker
matching it with the number on the package and both immediately alerting the user also and immediately alerting the backend also. Because half the time it's not actually fraud. Like the guy just literally picked the wrong package and kept to your door and you know, the package has got swapped between the neighbors.
00:46:21
Speaker
And then if if it's valid, then that proof of delivery is also authenticated. So it just reduces the back office job a lot. so and okay like Any place where we have basically unstructured data and in logistics, you have a lot of unstructured data, you know, the 2024 LLMs and AI systems are effectively helping it convert into structure.
00:46:42
Speaker
Fascinating. ah yeah I guess the the two big pain areas for D2C brands are... ah the RTO, which is like when an order goes back and the cash on delivery.
00:46:55
Speaker
ah Do you support in solving these pain areas? Yeah, and we absolutely. So we do both cash on deliveries and return optimizations, right?
00:47:06
Speaker
For return optimizations, one of the coolest things we did is for this Apparel brand and they they offer three hour returns. Now on the surface, it would look like an insane cost.
00:47:20
Speaker
Three hour return, right? You raise ah return request during during working hours, not 24 seven, right? like The trick is we actually, so let's say the delivery starts from 8 a.m., right?
00:47:35
Speaker
The returns are only started from 10 a.m., right? So between 8 to 10 a.m., all your delivery executives now have some latent capacity. This company is only doing apparel delivery. So the transaction cost of delivering picking up is very little.
00:47:52
Speaker
So now we find the nearest driver who both has some latent capacity in the bag and also latent time, right? Some time slack to meet the next order.
00:48:05
Speaker
And we add this return request there. So your total landed cost barely increased. And what you could offer is effectively a service where customers can order a lot more stuff at far fewer friction to cart conversion and easily return it.
00:48:21
Speaker
And just that you know confidence a customer gets by being able to return something today and today, it's amazing. um Right, this try before you buy kind of... a Yeah, yeah. So the brand, you know, brand managers mentioned about, you know, how people wear the clothes they bought only after the stuff they didn't want get got returned.
00:48:42
Speaker
So this very clearly, you know, some safety security mechanism happening in your mind. Right. So that's what that was on the RTO piece. the What was the other thing you mentioned on the cash on delivery?
00:48:55
Speaker
Cash on delivery. yeah Like brands want to reduce cash on delivery, right? They want more orders to be prepared. Do you support in that? Like. So we would not be a helpful solution for somebody wanting more orders to be prepaid. That falls outside of our purview.
00:49:14
Speaker
We handle the cash on delivery orders. And not just D2C, one of our big customers, Big Basket also has cash on delivery. So, yeah, so we would do the cash on deliveries. We would help you reconcile the cash at the end.
00:49:26
Speaker
If you have like a disbursement system, we will link it with that as well. So yeah that, you know, after end of the day, when the driver has already spent the entire day delivering orders, he's not, you know, waiting half an hour in line to sort out his cash.
00:49:39
Speaker
Right. And just that adds a lot of happiness to them. Right. How you end your day impacts how you think about your day. right So there's there's a head of supply chain at Big Basket, head of supply chain, KB and Mr. Nagaraju.
00:49:54
Speaker
And he came up with this idea and that's how we sing this all together. And the driver happiness increased tremendously. The ah problem with cash and delivery often is that those orders are often frivolous.
00:50:06
Speaker
Like the, I mean, the delivery actually doesn't happen and the executive ends up having to take it back or or whatever. so So do you, I think with UPI now with cash and delivery, we are seeing at least in our transaction volume is slowly getting to a point where brand managers are choosing not to go with that.
00:50:26
Speaker
At least in the urban centers. As in they have stopped offering this on the checkout page. yeah the yeah the Or even if they're offering the volumes are minimal in the eight cities is in India.
00:50:38
Speaker
Ah, okay. Okay. Okay. You would actually now in Bangalore see shops which says no cash, only UPI. this is I saw it for the first time in my last visit.
00:50:50
Speaker
This is actually post-COVID extremely common in UK, which is like card-only systems. But yeah, now we're starting to see a bit of that. Because, you know, even for ah like a small shopkeeper, just managing that change is not... and And there's a lot of pilferage in it.
00:51:07
Speaker
Right, right, right. Okay. So ah tell me about your... ah the go-to-market journey, you said your first order was $250,000? No, that was that was my first My first customer also, my first customer, like yeah, yeah, lack of right, would be still under a year.
00:51:31
Speaker
okay and like so definite three hundred thousand dollars a month ah right at start and ten thousand dollars yeah under ah yeah and Today, our biggest customer would be a few million dollars a year.
00:51:47
Speaker
well yeah if loy and And these are often like multi-year logged in deals. So these are often like eight digit dollar deals. So how did you, like, ah ah tell me that journey of, you know, your average contract value from $10,000 to multi-million dollar average contract value.
00:52:07
Speaker
How did that happen? What worked? what What was the way in which you traversed?
00:52:15
Speaker
Yeah, I think slowly and steadily, but ferociously. okay This is not mine. I think this is the motto of pure origin, granitum. I don't know to pronounce it in Latin.
00:52:30
Speaker
But yeah, so I think one of the first things we did was take a call that we're going to be an enterprise company. And this happened about 2018. So about somewhere in the three to four year mark. This happened with our work with Unilever.
00:52:44
Speaker
And with that, we started first changing our tech stack. right so we needed to get So we were one of the first, one of the youngest companies in this space to get ISO certified.
00:52:55
Speaker
right security standard Now there's no GDPR certification, but our security and privacy standards far exceed them. So we are now a GDPR++ plus plus compliant. Same with British privacy standards.
00:53:08
Speaker
So first step was ensuring that your product had the right components for an enterprise to even buy it. then in then and Then the second part was, as you move towards these high order values, you have to introduce new buyers.
00:53:25
Speaker
So when you are selling something at $10,000 a year, there's no procurement team. You're selling to the business guy and that guy signs it off. When you start selling it at $100,000, you have to convince the business and that business guy will champion it to the you know exec and you will have to have that group.
00:53:43
Speaker
Group buy-in, right? When you start selling it at a million dollars, you need to start convincing the IT team that, you know, this would be the change management and that's how it happened. You need to convert the CFO office that this is the right yield.
00:53:55
Speaker
You need to convince the procurement team that, hey, you know, we are a stable organization. We got to remain in the business. This is a competitive rate. And then, of course, you got to convince the business and product of your whole value. But, you know, in hindsight, it seems obvious.
00:54:07
Speaker
ah But, you know, when i was 26, when I started this, had no idea. had never done an enterprise sale. I was this clean-shaven guy, always called beta, beta, beta. yeah So yeah, these are all, you know, stumblings and learnings.
00:54:21
Speaker
Like in in in my early days, I still like over like, you know, 200 domestic flights a year. would literally just be always in front of the customer. And that's how we picked up these things. What, you know, so if you were to hire someone to hypothetically, you you decided to hire someone fresh out of college and make him responsible for sales, ah what would you teach him?
00:54:45
Speaker
Like in terms of how to sell?
00:54:50
Speaker
um how to and okay How to get those multi-million dollar deals. Yeah, yeah, yeah. Okay, so that changed. Okay, let me break it into two, three parts, right? Just first on the sell, then selling to an enterprise, and then on the multi-million piece, right?
00:55:10
Speaker
First on selling, and this is, you know, not just to a fresher, this is something we constantly keep correcting ourselves, right? After every meeting, and you know, we do. but because you know fundamentally we're engineers so we love our features we love the specifications we have built we love what it can do we love this one more configuration option right but during selling you got to love the benefit your product will bring to the customer that the customer wants your product may be teaching buddhism but does the customer want it so building but you know and
00:55:48
Speaker
And it takes, it takes, it's just not, you know, words. You got to understand the customer's business. You got to have the right research. You got to understand the individual, right? So for example, now when we approach public market companies, we scan through their financial reports of last few years.
00:56:03
Speaker
We understand what were the goals that they had talked about, right? What, what, what is their terminology? For example, if you're selling to a pharma company, right I think you should not call it drivers, you should call it flipotomist or field executives.
00:56:17
Speaker
So how much you can relate to the thing. So I've now segued from selling to selling into the enterprise piece where I think understanding the various motivation of various buyers and the various parts goes into the multimillion thing that my the value of my product to the head of the IT is very different from the value of my product to the head of sales.
00:56:42
Speaker
And both are equally right. So for the head of the IT, I talk about you know how we have previously engineered AWS systems and our API uphold the similar levels of public standards, how you will get callback APIs and how you will not need to constantly pull us,
00:56:57
Speaker
we will push back updates, how it will be less expensive, how it will be easier, how the integrations will be robust, you know, how it will happen right the first time, right? And that's a completely different messaging to what I would do to, say, a CFO, where I would often go with a DCF or an NPV analysis that, you know, hey, so give me if we are yes discounted cash flow and net present value, right?
00:57:20
Speaker
So we are often the most expensive solution between the competitive landscape. So we... at the start, right? And that's what we tell ah tell our buyers that, you know, look at all our historical buyers, talk to them that we are the lowest total cost of ownership, right?
00:57:37
Speaker
Because we are saving you money, right? So often how we even bill our product is effectively, you know, 10 to 20% of what we expect to save you. And that has led us to, you know, tremendously grow in our accounts as well while cracking these large deals, right?
00:57:55
Speaker
Third, just specifically on the topic of, you know, very large enterprise deals, right? Often conventional wisdom, you know, and stories and anecdotes had told us to appear bigger than you are, right?
00:58:14
Speaker
When cracking these very large deals.
00:58:18
Speaker
In my experience, I've found it to be opposite, right? I think you can do that yeah Also, yeah, it's just not the most integral thing to do. But also the buyer, right? You see, if I'm talking at the highest end of your spectrum, you're ah clearly punching above your weight class.
00:58:35
Speaker
You will need to find a buyer who's keen to work with a slightly smaller company. Who sees value in why this younger company who's more agile and has more recent tech will actually solve their problem.
00:58:51
Speaker
right So this is my key advice that when selling your largest possible deal, please sell it to the person who knows everything about you and wants to buy.
00:59:04
Speaker
Because this is a start of a five year relationship. Right. And that person will always have power over you because that person is giving large parts of your cash flow. Right. But if then that person is somebody who understands that you're the younger, these are the advantages they get. These are the disadvantages.
00:59:19
Speaker
They can actually really help shape you. You can, you you can. So for example, when COVID hit, I'm so grateful that two separate customers of mine called me and said, Hey, Nishat, if you need money, we will offer it at zero percent interest rate of our balance sheet, but you can't go down.
00:59:38
Speaker
Right. And that's what these legit customers and we find them both in India, we find them both in international. And of course, at that scale, everyone is a public market company. Right. We're talking at the levels of Fortune 100 companies, not even Fortune 500.
00:59:51
Speaker
And I think today the best companies, the best growth companies are eager to partner with, you know, young, young companies who can solve a specific problem to them, have better tech and are humble about it.
01:00:04
Speaker
and So like be authentic is like like one one big thing to strive for in sales conversations. ah Especially on your largest deal, because it can go very wrong if you don't do that, right? Like don't put up like all red flags put up first, right? So to all my largest deals, you know, know I would have almost opened with this way that, you know, just I'm so thankful for you for presenting me an opportunity to replace a software that was my ideal when I started.
01:00:34
Speaker
right Because all of these big companies, not like the first time they're buying the software. i'd like They have these legacy TMSs that, you know I mean, you know Oracle SAPs of the world is what we grew up idealizing.
01:00:45
Speaker
And now we have become a legit competition to them. And it feels great about that. And I think I'm very thankful to my largest customers for making that happen. By the time COVID hit, what kind of scale were you at? What kind of average ACVs were you doing? What was your ARR?
01:01:02
Speaker
So when COVID hit, we were at a few hundred K. Post COVID, we have now scaled to like a million dollar plus and with the largest ones being even higher.
01:01:14
Speaker
And what was your error when COVID hit?
01:01:18
Speaker
um It was in low double digits then. Right. Million dollars. Yes. And what is it now? yeah It's scaled. So we've been fortunate.
01:01:30
Speaker
Three days it? So, you know, what was tough was the 22-23, where basically everyone imagined home deliveries to have completely shot through because of COVID, which was true, but it was true only for the existing players, right? 23 middle was the quarter when we had, when globally the least amount of software was sold.
01:01:53
Speaker
Why is that? Because you know everyone was working from home and all, and it takes and time for enterprises to come back. Then they attack the mission critical things. right So Q3 23 globally was the lowest enterprise buying activity.
01:02:07
Speaker
sit But that was also the time, you know, most of the ah RFIs were coming, which got converted into ah RFPs and now which have converted into great closures. So, yeah, I think the whole, you know, contradiction about you expecting a lot of sales to happen and not happening was a bit tough, but both pre-COVID and post-20 and 24 onwards, it's been a great journey.
01:02:29
Speaker
Are you 50 million plus today? No. Okay. Okay. Okay. but But by when do you think you'll be 50 million plus? like Over the next or a couple of years.
01:02:42
Speaker
Okay. Amazing. Amazing. Okay. So, ah you know, what is the, you spoke of the legacy competition you have in the form of an Oracle and an SAP.
01:02:53
Speaker
ah They also have like root optimization, right? ah So, for example, many of them had it at a level of PIN code, right? Today, e-commerce operates with like three distribution centers within a PIN code.
01:03:06
Speaker
Right. They had it as allocation over, you know, optimization, right? And it was still designed for, you know, full truck loads moving between cities on Monday, Wednesday, Friday, right?
01:03:19
Speaker
Not like, you know, one vehicle doing 60 deliveries in a day in an urban environment. and So who are your modern compet competitors? And yeah, and what that has changed is that has also changed, by the way, your long haul deliveries.
01:03:33
Speaker
Now B2B operates on this thing called B2B Express, right? Where 24 hour deliveries happen because same, your B2B shopkeeper is also an Amazon user, right? So now he's asking Unilever, why can't you replenish my inventory?
01:03:46
Speaker
every day, right? I have the same shelf space. Now you have six different kinds of Vaseline. Previously, you only one Vaseline. How will I stock all of this inventory? You need to replenish it. So there are actually, you know, stores in Tokyo, which because the real estate is so costly, where replenishment happens often three to four times a day.
01:04:05
Speaker
Wow. Okay. ah ah Right. the Who are the modern competitors that you have? Sure. So so ah we have Brink,
01:04:17
Speaker
in Tel Aviv as a US team, right, that operates in the North American markets. Then from India, we have Farai as a competitor. um We don't see them much in the North American side, but we see them a lot on the India, Middle Eastern side, right?
01:04:33
Speaker
Previously, we used to have Loginex, but over the last two, three years, they haven't been that active, right? Yeah, so that would be on those new age companies. The rest of them, I think, attack much smaller companies and not in our space.
01:04:46
Speaker
Okay, okay, okay. And ah what is your global split of revenues? Like how much of your revenue comes from India? How much from US and other markets? and Yeah, yeah, yeah.
01:04:57
Speaker
So about 60% of our revenue comes from India plus Asia. So we also have teams in Indonesia, Vietnam and Thailand. Those are important reasons for us. So we actually first did near shore before we went far shore.
01:05:10
Speaker
And about 40% of our revenue comes from the Western markets. Okay. Okay. Okay. And I'm assuming- There was small sliver in the Middle East as well. But yeah, that's not a big region.
01:05:23
Speaker
West would be the focus for growth, right? Because much higher HDV is there. and So how how are you how did you ah go global? Like, you know, moving out of India, how did you learn to sell beyond India? Yeah. um I think we made all the mistakes. That's how we learned how to sell. But yeah, fundamentally, we started first with near-show strategy. Why? Our thing was to bake our product, right?
01:05:51
Speaker
If you design a consumer good for a person in California, it works everywhere globally, right? But if I design a routing system for, you know, California and Berlin, it will blow up like a high school project in most of the Asia, right?
01:06:08
Speaker
So we wanted to attack the toughest geographies for the software first and then move upmarket. This also makes your product very robust. Transport management systems are also systems, you know, where which have a shelf life of 20, 30, 40 years, right?
01:06:23
Speaker
So it takes it's it's important to do it right. Like often, you know, in many of the deals that we are in, they they have this policy that, you know, they will just not engage with a sub five-year-old company, right?
01:06:35
Speaker
So they don't want a 20-year-old company, but they definitely want like, you know, a profitable growing or a cash-rich company to work with because these are very, very long-term engagements, right? So we started first by Nearshot, that allowed.
01:06:47
Speaker
But that led us to regions where we don't speak the local language and we had to, you know, convince a lot of the local, you know, So we we hired locally, right? It was initially a lot of founder led sales, but we had like, you know, some young guys in the team at 24, right?
01:07:04
Speaker
And one of them is now our chief product officer, Pratiman. He actually went and opened up the entire region, you know, hired locally and got live the first few clients. How do we get the first few clients?
01:07:16
Speaker
Thankfully, India and Asia and now even US s has a lot of common investor circle. So we first started approaching, you know, portfolio companies of our investors. Or if you were working with a company in India who either had a sister company or they had operations globally, that gave us like the first customer there based on which we could then build referenceability.
01:07:35
Speaker
Word of mouth still remains like a main GTM channel for us. Okay. Okay. Because logistics is a, you know, industry where people often move companies within. So happy users are great.
01:07:46
Speaker
We have like internal systems to track our users and how they're moving across companies. And then we just reach out to them. We love putting our product first. Okay.
01:07:58
Speaker
How do you, uh, like, you know, you, You mentioned to me you had like a quarterly business review with a client before we started recording. ah What are the processes ah and mechanisms to ensure that there is no churn, that ah you are able to get more revenue out of existing clients? what What are some of those systems that you have set up to drive that?
01:08:24
Speaker
yeah Yeah, so we have set up systems and s SOPs around it, right? And they they involve doing QBRs and things. But our fundamental tenet is to be paranoid about on-ground value, right?
01:08:40
Speaker
So for example, let's say if we have a contract where I've been paid for one year upfront, right? And in the third month, I'm doing a review with the team and they mention, yeah, we are getting paid.
01:08:51
Speaker
That is completely irrelevant to us. That is the CFO's problem. That is not my problem. right My problem as the product person is to ins ensure not just they have used, but what is the value they have derived in the graph.
01:09:09
Speaker
So let's say if there's an adoption issue that all technical integrations and all are done, but the users have not adopted. right So we will obsess about it. We will go to the users. We'll ask them if it's a bad user. we We'll tell the warehouse manager. If it's often the users have some genuine issues, we'll incorporate it back into the product.
01:09:26
Speaker
But fundamentally, if your product is working on the ground and adding value, right, even if your customers are complaining or raising, that's great. you know customers who don't say anything are actually the customers at churn, right? All engaged customers are good.
01:09:45
Speaker
See, we have to take the accountability that we are a critical path system, right? If we go down, it won't be an inconvenience for anyone. Their business will stop.
01:09:56
Speaker
Their CEO will know in next 30 minutes on the internet, right? So we have to operate with that level of accountability on our end as well. right And yeah, so while we have these QBRs, SOPs, those reviews to make sure, but the one thing, the one fundamental tenet we have with every technical account manager, every customer success manager, every engineering, right?
01:10:17
Speaker
Are we adding value to this client? How do you track that? What metrics do you track? to Thankfully, that's a very objective number. right So you can track value in several. So there is there's a lot of intangible value, but thankfully there is a lot of tangible value also. You can just simply at the very basic level, you can track average number of kilometers per order or number of deliveries done per day by per driver.
01:10:44
Speaker
That's a metric your driver understand. That's a metric your CEO understands. Time utilization on Right? Yeah. So there are very hard, tangible, measurable metrics, which you can account for. And that's where you know that, you know, if you're adding value, even if somebody is complaining about the user interface or a thing, like that's fine. Right? Like that's a, but this account is not at risk because you're adding value. Like, also and i told you ever undercut on price.
01:11:11
Speaker
Right? So we are anyway, never because you know, this guy was right. So it's always a pressure on us. Hmm. Right, right, right. And these metrics you will track over time to see how they have improved.
01:11:22
Speaker
Yeah, absolutely. we have We have respectfully, you know, bowed out from conversations where we have been pushed, key you know, so then that's not, you know, that's not what you're buying, right? You're not getting it. It's okay.
01:11:34
Speaker
Hmm. ah You said you typically try and charge 10-20% of the amount you're saving. ah how how do you estimate and what will be the savings? In large accounts, you will rarely go live without an on-ground pilot. right so there's ah There's an entire discovery phase that happened between two, three months.
01:11:51
Speaker
We get a lot of their data metrics. We simulate. right we will also That's why I tell them that these are the areas of first account. See, that's what at at a large and enterprise level, the buyers are extremely rational, right? They have a very real defined problem and they need someone to address, solve it.
01:12:09
Speaker
With a large enterprise, the sales itself would take maybe six months. It takes long. It takes long. Could be longer. Wow. Okay. Could be longer. Could be a year, right? ah If you're doing like an eight digit dollar deal, it can easily go over a year, right? that are and So while there is a lot of land and expand, as an organization, we equally believe in signed and expand and landing large.
01:12:35
Speaker
Right? Because, yeah. so and So, you know, in our first five years, we would often also, you know, just take like a small sliver, deliver and then expand.
01:12:46
Speaker
But now we are also equally focused on large on landing large and sign and expand that once we're into the organization, as soon as our integrations are starting, we'll keep asking them, you know, if there is a parallel country we can go to, is there another division we can deploy with, right?
01:13:02
Speaker
Over waiting till land and expand. There's a, um I you forget the name of the gentleman. He was the CRO of Annaplan. And he has a beautiful small book called Selling to the Cloud.
01:13:16
Speaker
It's a great book. And this is an idea we picked up from there on focusing on sign and expand over just la land and expand. What is the difference? Land and expand means once you've deployed and stabilized. land expand is like, so yeah, exactly. It stabilizes. You show the first value. right Now the thing is that often takes like six to eight months. right And then the customer, OK, do this, do that. right but you go to you But instead, like you know the moment we close the contract, that's the next meeting we do is like, you know who else can we talk to?
01:13:45
Speaker
right And they will we will obviously that hey do this then. But sure. But you've got to keep pushing through that.
01:13:52
Speaker
and then And then during the contract now, you know, yeah we would that, hey, okay, let's look at a multi-year, right? Like we'll give you a 10% discount if it's like multi-year, this scope, right? So yeah, we are also equally focused on landing large and over, we have realized that, you know, basically we can compress the sales cycle only beyond a certain point.
01:14:11
Speaker
So now I would rather take another month in the sales cycle, but, you know, lock myself in at a high digit value. Okay. Okay. Okay. ah I assume you would, the relationships with customers would be something that you would need to obsess a lot about, right? Because essentially, like if you've converted somebody who's handling transportation in one country, you need him to champion you to other countries. You need him to do introductions. How do you map your relationships with customers and like, you know, uh,
01:14:46
Speaker
probably be based on problem statements, right? So we we solve a problem statement for an individual and then they are you know without they are more than happy to champion us everywhere, right? like So we often we we generally engage with our cousin you know are stakeholders and ask that, hey, what are your KPIs, right? Like what will help you get promoted, right? And we first focus on that success.
01:15:10
Speaker
Many of our customers are are our customers in the next organization. Some of them have actually over time joined us. Many of their family members now work with us. Like they have been great referrals for you know potential hires.
01:15:25
Speaker
Right. ah So, yeah. And see, we were outsiders to supply chain. Right. We very clearly used to tell our client that that's why that's also another reason we have a non-opiniated system that we know maths.
01:15:36
Speaker
Right. And what we were going to give you is a system which is elegant enough to do configurations over customizations. Right. That we will actually take, we take. So, for example, we have practically built our own graph language to model any kind of, you know, distribution system onto us very quickly.
01:15:55
Speaker
So we invest a lot in the backend infrastructure that there is elegance in the system that you you'll be able to configure it for various use cases over having to customize it. And, you know, that slows the whole thing down.
01:16:06
Speaker
Right. We focus on those individual success and then we, you know, but at at first you got to deliver value.
01:16:16
Speaker
Configuration over customization. So essentially that means that you can, you Create as much opportunity for things to be not hard coded, but variables. Parameterized.
01:16:32
Speaker
Parameterized as a philosophy of how you build a product so that as much as possible can be configured and doesn't need any major. Now that does something very tremendous for our clients.
01:16:45
Speaker
It leads them to a product which is robust. Many of our clients at first time often ask us for some custom best spoke stuff, right? And I tell them, you know, you can do it, right? And our APIs are open and this is a, you know, SI team which will do it. We can build it transparently. I don't even want to make money on that.
01:17:06
Speaker
But if you use this with this small sacrifice, what you have is an interface which has been tested, you know, across hundreds of devices.
01:17:17
Speaker
written with the highest standards of human computer interaction, right? And tomorrow when your use case will change, it will still adapt. So sometimes obviously, you know, like 10% of the entire 100% scope in a large multimillion dollar deal, it doesn't matter, right? But fundamentally the core, right? So for example, we never write custom routing stuff for anyone.
01:17:39
Speaker
That's a like a hard no. We will focus why our routing can't do it. What is the elegance it is missing? And then we will go fix that engine itself. Okay. okay So having stuff which has aged well in in adds a lot of robust robustness. right so hence also elegance and focus on configuration over just every customization.
01:18:00
Speaker
What did you mean when you said you built your own graph language? What does that mean? Sure. Absolutely. So let's say, you know, ah first mile logistics, last mile logistics, right?
Tech Innovation in Logistics
01:18:12
Speaker
They are effectively inverse of each other.
01:18:17
Speaker
Right. If I'm picking. So if I have to think of the entire logistics everywhere, right? Think of it as ah vertices, right? Nodes and edges connecting them.
01:18:29
Speaker
edges have a certain relationship with the two endpoints and there are certain activities that happen at each endpoint. So when you are talking about cash on delivery overnight, so now if you model it as an activity on the edge, instead of a credit card swipe, I can have a cash on delivery.
01:18:45
Speaker
right So effectively, all the possible combinations of things, you know they they are like a graph. So we have also modeled it internally. So often, you know we we would even do in our early days things like you know if somebody drew it on a whiteboard, we'll take a picture, we'll send it out to our solutions engineering, and we will actually quickly get that same model life.
01:19:05
Speaker
And then we'll be able to simulate it So that's the idea. right like Invest in building that core infrastructure. and Okay. Okay. ah so but I'll tell you something live also. So right now, you know, now our product has really expanded into this and we noticed when we added our last module, it took us longer than we would have liked for it to start talking to everything and to integrate, right?
01:19:31
Speaker
And this means it is about time for us to start investing in ah slightly more sophisticated internal data lake, which can be accessed in real time and runs across.
01:19:41
Speaker
So now this year, maybe we're going to invest in building that core infrastructure, which in turn will make our future application development even more. Okay, fascinating.
01:19:52
Speaker
ah And this because you yourself are a techie and the the product guy, so you are able to ensure the highest amount of optimization in how the product is built.
01:20:06
Speaker
Yeah, I mean, and also have like a really great team around me. My co-founder, Geet, he's a great engineer who helps scale everything and make sure nothing ever breaks and has the highest level of security standards, right?
01:20:18
Speaker
So he's he's the person responsible for production engineering of everything. Mm-hmm. Why did you choose a tech co-founder instead of a business guy?
01:20:30
Speaker
Oh, I chose a co-founder or choosing a tech or a business co-founder. Okay. And also 100% of my circle that time involved people who code. Okay. Okay. i was like yeah I was writing a lot of open source core before I graduated and you know high school. And then there was college and that was an engineering college. And then I joined Amazon. That's again an engineering heaven.
01:20:55
Speaker
right And you know we just wanted to build great stuff and eventually find some monetization path. And yeah, so that wasn't that wasn't even a question, right? It was just like ah two two engineers trying to build something cool.
01:21:12
Speaker
So even today, like, as I said, like, you know, our defining motto is to invent the future with amazing people. So invent emphasizes on actually accomplishing it, not just thinking it.
01:21:23
Speaker
Future you know emphasizes on the excellence of the tech we build, on the deepness of the tech we build over you know any commoditized. And with emphasizes that you know it's a problem of scale. Two people can't do it, so you need other people.
01:21:38
Speaker
And amazing people. And how do i define how do we define amazing at Locus? Effectively, the three I's. Amazing people have great ideas. Ideas sit at the intersection of knowledge and insight.
01:21:52
Speaker
Ideas are not just candies, detailed thought through ideas. Second, they have great initiative. Initiative sits at the intersection of energy into ambition.
01:22:04
Speaker
that You should actually have that raw energy in it. You should have that both physical stamina, that mental stamina, right and then the ambition to apply it to toughest problems. And third, integrity.
01:22:17
Speaker
Integrity is the intersection of doing the right thing in tough situations when no one's watching.
01:22:24
Speaker
So that's how we define amazing people. And yeah, so our core mission always is to invent the future with amazing people, and hence the focus on technical excellence.
01:22:37
Speaker
Thankfully, we were all born in a decade where we could do this. I don't think this was possible in 60s. But I think it was it was the probable path in 2010s and 2020s. And now this will become the first path.
01:22:51
Speaker
Did this force you to become more of a business guy? I mean, yeah, absolutely. but if you i first but One of us had to write more code. One of us had to write more And I thoroughly enjoyed that. Right. For me, becoming a business guy was nothing but running a bunch of small A-B tests.
01:23:13
Speaker
right I took it as a very engineering problem. And even at sales, right? Like how we say it is that, you know, engineering is like a tightly coupled system. It's like a wristwatch, right? If I do something one, I know exactly seven degrees apart how it will happen.
01:23:30
Speaker
Sales and organizational behavior is more like the human body. It's a loosely coupled system, right? So if I do something seven degrees apart, I have confidence interval with a probability that these set of actions can happen.
01:23:43
Speaker
So if anything, you know, actually imagining a loosely coupled system is kind of like, you know, a tougher problem than a very tightly coupled system. And that's how we're all very excited also about solving business problems.
01:23:55
Speaker
yeah Do you, like like, is that how you think? ah Like, you know, if you decide, okay, I want to expand my US sales. So like, okay, if I do this, then seven degrees later, my sales could increase between 5 million 15 million.
01:24:10
Speaker
fifteen million To some extent, absolutely. Right. And you know how we we we try and do it at a very, very first principles level. So for example, you know, when people ask us our GTM, you know, and how we're deciding on GTM, I actually say a mathematical ratio that we go after industries with the highest cost of goods to cost of logistics ratio.
01:24:33
Speaker
For example, if you sell a candy, right, 40% of the cost of that good is just logistics. let's want the okay So if I optimize this and the profit margin is very thin, it's just 1-2%. So if I optimize the cost of logistics by 10%, 40% of 10% is 4%.
01:24:54
Speaker
That 4% is often more than the profit margin. So my buyer is massively incentivized to use a system like me. Contrast this to jewelry. Even if the absolute cost of moving a necklace is very high compared to the cost of the goods, it's nothing.
01:25:10
Speaker
Right. Even if I half your cost of logistics, you don't give a sh**. Like you don't care. Yeah. Okay. So how we approach our client, we like to be very first principles, very mathematical.
01:25:22
Speaker
i mean, that's our edge. There are other ways, you know, I'm sure there could be better ways, worse ways, but our edge is maths and first principles. And that's how we approach even business problems. ah Give me some other examples of where you use math for GTM, for sales, for ah organization building, for some of these like loosely coupled systems.
01:25:42
Speaker
Sure, sure. Or like some of these non intuitive ones. Okay, so like, we wrote a simple script in our early days to check for Indian companies where there is a CXO with the same surname as the founder of the company.
01:25:57
Speaker
But the CXO is under 40 years old and the company is over 40 years old, right?
Hiring and Team Building Philosophy
01:26:02
Speaker
That effectively gave us a list of c sons of sons and daughters of promoters who have finished their education, have joined in as some chief strategy or chief innovation role, and they want to you know do something big.
01:26:16
Speaker
They also have the right say in the organization. And then we first started doing GTA. And those are also people you know we could relate a lot more to go to the same cafes or same bars, have a conversation, right?
01:26:28
Speaker
We did the same with hiring, cross-referenced it. you know Initially, when we didn't have a lot of money, so instead of going after you know kids from IoT and BITS at the very early, yeah and we wanted to get great engineers, we actually cross-referenced GitHub contributions right to like the second layer of colleges.
01:26:46
Speaker
and referenced it to certain social media posts. So that's how we zoomed in on some of our talent finding. Yeah. Wow. Amazing. Amazing. ah how do you How do you screen people when when you're hiring talent? What's your screening method?
01:27:04
Speaker
Yeah, I don't think I can sum it up to one. I was thinking if I have like a you know primary screen at each level, right? And I can definitely think of examples I use to reject people.
01:27:15
Speaker
but there is definitely not a single screening to accept people. And that both changes by the seniority and the roles. Right. What I mean by, you know, methods.
01:27:26
Speaker
So for example, you know, if in an interview, say somebody is from a very pedigreed college, right. Or let's say I'm from bits and they are from bits and they talk a lot about bits and the past that doesn't really board well with us.
01:27:40
Speaker
Right. Like, uh, we want We want to hire people who have, right? We are looking for exactness in your conversations, right? If there is a lot of ambiguity, vagueness, collectivism, right? That we are also right now biased towards more doers over managers.
01:28:03
Speaker
Integrity is then becomes a very, very major screen. right So we'll put various case studies that, you know hey, you're in Vietnam. This is this the cost that has gone in this year. What happens?
01:28:15
Speaker
Because you know for many, you know their workplace is their temple. right For me, my workplace is my religion. It's my god. It's my temple. Both the founders and many of the early team members you know Like they have sacrificed large parts of their personal life to make this happen.
01:28:34
Speaker
And at this point, one of the key screening that remains is that everyone we help join into the organization holds integrity at a very, very high level, right?
01:28:45
Speaker
So yeah, we don't ah just academically, we try and screen it in questions, right? But fundamentally, as I just mentioned prior to this, what we are looking is are for people with great ideas, great initiative and great integrity.
01:29:00
Speaker
Those are our three criteria. yeah okay yeah ah It seems like you're looking for an engineering mindset. like They may not be engineers, but that engineering mindset of ah systems thinking, of applying math. Problem solving.
01:29:15
Speaker
Problem solving. Yeah. Yeah. Yeah. Which is, which is even more critical now, because if now you can actually define your problem and structure it into steps, as we were referencing earlier, several LLM tools become great craftsmen for you to actually execute it.
01:29:32
Speaker
Right. Yeah. Yeah. Yeah. true So now you need ah great ideator over a team of content writers. True, true, true. Yeah. Yeah. Your team is the LLM.
01:29:44
Speaker
right so yeah so uh you have raised more than 80 million dollars till date uh what have you learned about fundraising and you know both in terms of how to raise funds when to raise funds should you even raise the uh funds to the extent which you have um
01:30:08
Speaker
Awesome. I would also add one more, you know, in that when and how one more dimension, which I think is even more important is who to raise funds from. Okay.
01:30:20
Speaker
Right. And while that evolves through your fundraising journey, but my singular advice to everyone would be to always keep that as your tier one focus of who you are raising funds from.
01:30:34
Speaker
And I mean, which organization and then which individual within that organization, right? I say this as a learning from complete positivity.
01:30:46
Speaker
We have fortunately been blessed with investors who during our tough times and during our good times were there equally were there equally supportive. And that made not just surviving through the tough times, but to focus on the right things like the team and product so much easier.
01:31:04
Speaker
So if you are if you are looking to build a company for a long time, you have to understand that an average investor-founder relationship may outlast an average marriage, right?
01:31:17
Speaker
So it's really important who you raise funds from.
01:31:23
Speaker
And that is where you know a lot of institutionalized seed and venture capital, right? Like why not take a debt from the bank, right? Why not, right? It's important to raise funds from people who have, you know, built companies or seen companies, seen multiple companies, you know, go from zero to 100.
01:31:41
Speaker
They have the right guidance. And also in the most critical moments, they have the right compassion.
01:31:50
Speaker
So yeah, I would strongly focus as much on who to raise funds from. Okay. Then comes on when to raise funds from. Right. Uh, very, very, very early on somebody had told me this one line, which kind of really, really stuck with me, which was very nice.
01:32:09
Speaker
yeah very nice Uh, yeah.
Fundraising and Capital Strategy
01:32:16
Speaker
So I think, uh, we only did demos for our fundraising very oddly.
01:32:21
Speaker
Okay. Okay. Every, every time, you know, the few, and you know, as I mentioned, I, we keep looking at our AB test, right? Every time we started with a pitch, it didn't, you know, result in the fastest conversation.
01:32:33
Speaker
But every time there was either an investor who has a portfolio company where we have already delivered value. And as I was mentioning one thing, you know, sometimes for a customer reference call, now we can be like, these are our customers go call them.
01:32:45
Speaker
Right. Like you're right. So either you had delivered value or we had started with a demo and vision.
01:32:52
Speaker
Right. And people had bought into that. product into that vision that, okay, this is cool. Right. Then are there any tricks of the trade? Of course.
01:33:03
Speaker
So I'll give you one, which, you know, which is always to, you know, somehow make your investors experience your product. So for my seed round where we needed the most credibility and, you know, nobody else was using the thing.
01:33:14
Speaker
I would always go and meet the investor over asking the investor to come to, right? First, I will go meet the investor and this is 2014, right? And when I'm going to meet the investor, I'll send them my own tracking link of RightSafe.
01:33:27
Speaker
right So now when they're waiting for me, they can actually see me coming. They've experienced the product. They've experienced the value of the product, right? It has actually solved, you know, some anxiety of them of when we are coming and I'll try and meet them outside the office. So they also have to come somewhere, right? So we'd meet in a mall, in a restaurant, and always just send them a tracking link, right?
01:33:45
Speaker
And 100% of those conversations after that was, you know, just about discussing the instruments of the deal, but I know like they were sold. Amazing. Amazing. Okay. Right. Then to address your third piece, how much to raise and think, I would say that, you know, raise a little more than you need.
01:34:06
Speaker
Not much more than you need, right? One, it's risk capital. Second, it always takes you a little more than you think. Third, if you have excess capital, there are always legit growth areas that you can build immediate adjacencies to, right? So I think...
01:34:23
Speaker
ah raising extra capital isn't the bad thing. Spending that extra capital without an abandon may be a bad thing. Okay. Okay. Your last round of $50 million sounds excessive.
01:34:37
Speaker
We still have a large part of it, by the way. Right, right, right. I mean, that that sounds excessive to me. Like, was it the post-COVID bubble where valuations were, like, ah really high? actually, our round, for what it's worth, you know,
01:34:54
Speaker
would be on the smaller end during those rounds that were happening. As a matter of fact, we were also offered two X of the money we raised. right We decided to raise half.
01:35:07
Speaker
right um And I think we have utilized this capital well. And we are now also investing in very interesting growth areas. So yeah, I'm very fortunate for having the opportunity to raise that capital and having great investors to raise that from, right? Like often what happens is when you raise extra capital, either you have to compromise on the quality of investors. We ended up raising from, you know, Qualcomm and GIC, like the sovereign funds, right?
01:35:34
Speaker
Or you have to optimize on certain or you have to sacrifice on certain terms internally, like liquidation preferences and all. Our investors were very graceful and, you know, it's actually great terms. And as part of this round, they actually even offered us the board majority.
01:35:50
Speaker
So, yeah, I thought I think, you know, with the benefit of hindsight, it was a great round done at the right quantum of capital. And it it allowed us the right cushion to go through COVID and not just survive, but thrive out of that.
01:36:04
Speaker
One thing that we did do was be cautious with that money. right ah Like most of our peers ballooned to somewhere between 1000 to 2000 people during those years.
01:36:15
Speaker
right We actually never crossed 300. Wow. Though ah this does put pressure on you, right? I mean, the more you raise the more... sort that pressure.
01:36:29
Speaker
I think it's it's very unfair to say that investors put pressure on you. right? You go to investors to raise capitals to reach your ambitions, right? You seek that pressures.
01:36:40
Speaker
Investors bring in accountability. I don't think they bring in pressure. Okay. I know of very few investors who say when they signed up on a deal and it was at say at X growth rate without any macro, without any significant change or significant change in capital and would randomly come and say do 3X without spending more.
01:36:59
Speaker
They would try and hold you to X and that's what brings accountability. And that's how you graduate into a professional CEO. And I think that's a very good thing. like Like you have to give projections and they hold you accountable to the projections
Leadership Transition: Founder to CEO
01:37:12
Speaker
you gave them.
01:37:12
Speaker
absolutely. Right? Right. The projections are in your... And that was a critical journey for us to, you know, transform. Personally, I would say from a founder to a CEO. And while I did stumble originally, i'm very grateful to have that opportunity and that time to get better at it.
01:37:30
Speaker
How did you stumble? so our projections used to be off. Simple. Okay. and Like you were over the optimistic. Yeah. And also see, I was a sports captain, right? Like ah that's how I started. I knew how to,
01:37:43
Speaker
lead 10 members, seven members, sports teams, and you know ah run it with that aggression. right And it was a journey for me to you know transform. and And I have four stakeholders. Investors is one of them.
01:37:59
Speaker
Team, clients, investors, media. right And how I interacted with each of them and primarily with my team definitely transformed as I took that journey from being a founder to a CEO.
01:38:10
Speaker
like Like you were overly aggressive initially before you realized that the larger the organization... yeah Yeah, right. Because like initially you operate as a as a sports captain and then you learn to recognize that you can, I forget who said it, but you you can never change someone's mind while angry.
01:38:29
Speaker
Right. And yeah, I think also as the years went by, i personally got the maturity of, you know, focusing on winning the war over winning the battle.
01:38:41
Speaker
What is the difference between the two? Like how, how, how, you can shout at someone in a meeting, but right? Prove to everyone that you are the smartest or whatever. Right.
01:38:51
Speaker
But did it accomplish anything with the next time that happened, not happened. Right. But if you could take that aggression, even if it is called out then, but follow it up with making sure that there's actual change in SOP or you write the email differently, then it results into action. And then you have actually won what has happened.
01:39:11
Speaker
Then you have made a meaningful change. ah What kind of a boss are you to work with? What what would your subordinates be saying about you?
01:39:22
Speaker
I would say they would call me a fair person. Right. I definitely have high expectations. Right. But we have, we have created a company for 2% of the folks who also want to have an accelerated career paths.
01:39:35
Speaker
But yeah, I'm very confident my team would call me a very fair manager, bias, hopefully towards excessive rewards, but yeah. Okay. Nice. And I personally, constantly, the one thing that I personally try and do is lead primarily by example, right?
01:39:54
Speaker
So, yeah, if I'm not able to solve a problem statement for my team, I know I will not earn their trust, right? So I believe everyone needs to earn their designation daily. Okay.
01:40:07
Speaker
At some point, you know, I used to say that, you know, for example, you know, back in 2017, you would have, you know, I would also say that hopefully they think of me as a great engineer or great this, right? But now we have evolved to a point where all my direct reports are much better functionally at what they do than i do My, I can't coach them functionally.
01:40:28
Speaker
My job is to empower them and make sure they're not blocked.
01:40:34
Speaker
And identify attribute to the right root causes. And that's where the fairness comes from. youre You're more of the architect now, like designing the organization and... ah I'm the enabler, right? Like if if, for example, product and engineering over a feature set is, i'm the I'm the one who would arbitrate it, make sure both see the same point. And that same happens between, say, product and sales, right?
01:41:00
Speaker
As we were talking about disagree and commit, my primary responsibility is to get people to commit on the same thing and say see see each other's side. That's the advantage I have, that I get to sit on data sets and interact with both sides of the equation and allowing me to build empathy on both sides.
01:41:18
Speaker
And then can I transfer and educate one side about the problems of the others? There were two ways to monetize what you have built. ah One is your way of SaaS and the other way is of taking a share from transaction, which I would say is like the Shiprocket way.
01:41:37
Speaker
um From what I understand, Shiprocket is in operations. So Shiprocket has physical warehouses and all. Right. yes Yes, yes, yes, yes, yes. You're right. Yeah.
01:41:49
Speaker
ah We don't do physical ops. Although like, let's say asset light ship rocket, one may not necessarily need to own the warehouses, but that could be other way, like the transaction way.
01:42:05
Speaker
No, that's still, so see, it's about who is owning the responsibility of operations, whether you're asset light or asset heavy, that's the operating model.
Tech Focus over Physical Logistics
01:42:14
Speaker
Okay. Right. So, for example, I only like my revenue to my customer is the revenue that is coming to me.
01:42:22
Speaker
It is not the entire cost of the delivery that I then take from. Right. Yeah. What you have built with minor changes could be the ship rocket route because you're essentially. just made a choice to not be in operations. Absolutely.
01:42:41
Speaker
we we have We actually got like several financial offers, you know, that, hey, why don't you start like an operations arm, start logistics. And especially, you know, during the cost of capital between 18 to 21, we had several options to, you know, ingest more capital and launch a logistics company.
01:42:58
Speaker
ah So why didn't you? Just the DNA of the team. Okay. Just the DNA of the team, sir. like ah I think we are much better managing technology than we are at managing other humans.
01:43:14
Speaker
We fundamentally at the core... It's still an engineering problem, right? Like it's just about tweaking incentives and managing execution. ums and Absolutely. Absolutely, right we We still think our edge lies in understanding even more complex technology.
01:43:33
Speaker
and adapting and simplifying it to the use cases. but So we love understanding the use cases, but the entire DNA of the team has always been towards focus more and more around technology side of things.
01:43:47
Speaker
I mean, Amazon is a technology company or an operations company? What would you call it? Amazon is a trillion dollar company. These things, these things, you know, ah dilute away at that scale.
01:44:02
Speaker
Right. true I mean, you, you, you know, you're not going to call Google a cap company, right? Google is a technology company just because they run Wimo. They're not a cap company. Right. Right. Right. So that's been our DNA and focus.
01:44:16
Speaker
No, I mean, just that the, the total addressable market is so much larger. The moment you get into the, the transaction sort of just- It becomes a different market, right? Then scaling internationally is much more complicated than say scaling with a technology solution. So today we operate in at least 20 plus different countries, right? And within them, the clients effectively operate operations in 30 plus countries.
01:44:42
Speaker
And that gives us a very good global scale, global problem statements. And that's what excites the team. That's what we are committed to ourselves, to our investors and our team member, that we shall focus on being a on a great technology company in and phase one in the transport space.
01:45:02
Speaker
Okay. Okay. Okay. Understood. Okay. So today, for example, you know, as we are gaining more and more retail customers and we are building, you know, we are getting into the adjacencies of pre-cart conversion, right?
01:45:14
Speaker
we We may end up designing a great technology company verticalized further into retail. So that's how we look at expanding our TAMP. But I think we will stick with our first love of tech.
01:45:26
Speaker
Are you planning to ah go down market like more SMEs or are you planning to just continue to focus on enterprise customers? and we have Actually, yeah, we we have been going up market over the years and we got we plan to stay on that trajectory.
01:45:41
Speaker
Okay. Okay. Okay. Also in SMEs now, I expect a lot of build versus buy pressures also, right? Especially also as the cost of development of basic systems have gone down.
01:45:53
Speaker
Ah, okay. ah And this is the impact of AI, like through you can- Yes. Cost of commoditized development has gone down, right?
01:46:04
Speaker
so if So what ChatGPT can't do today is produce and but use an algorithmic code like a routing engine, right? Or even close to anything to that. like So I want to go to customers for whom that is a core differentiating factor. And then I can charge even more because then for me also cost of other development is lower and I can give you like a wide variety of applications.
01:46:26
Speaker
How is ah AI, ah how are you using AI? Is it only for your internal product building? Oh, is it i no, absolutely. So, you know, as as a matter of fact, Locus did a public conversation about ChatGPT3, not GPT 3.5.
01:46:45
Speaker
And that was actually back all the way in 2021.
01:46:49
Speaker
Greg Brockman had to, you know, personally had given me an access key then. So we had been really experimenting and adopting this through and throughout. And for the first year, which was 2024, right, our goal core goal was to first adopt it in the organization before even in the product.
01:47:10
Speaker
So what are examples of that? Adopting. yeah and then And then in the product, it is both embellishing at each level. I was giving you example of that proof of delivery where it is, but also building an entire agentic TMS itself, which is the big.
01:47:24
Speaker
So now I will cover each of these three. First is say adopting it through and throughout the organization. right Let's talk about you know your people and culture. You can always write a slightly better hiring post.
01:47:39
Speaker
Right. Now, now something like a chart GPT can help you craft that. right Same goes for marketing, messaging, communications, design, right?
01:47:51
Speaker
Then there is of course engineering development, right? um Sales demos, right? Previously doing custom demos would take a lot of effort.
01:48:02
Speaker
But now because you know you can create like commoditized pages very quickly, you can have a lot of custom demo which is not linked to your product, right? So let's say if you're building a check, you you want to show how it will work with a checkout page, I can just, you know, today recreate the checkout page of the company I'm pitching, which previously had to go through engineers.
01:48:20
Speaker
So our first thing was, you know if I could share my screen, you will actually see that just two days back, we have said you know to the guys who lead like the internal change, have ah we have asked them to educate all the execs on how to make an LLM calls.
01:48:36
Speaker
Because look at, for example, my finance team. So, you know, my my CFO was visiting me last week because we were doing the annual planning. And we realized that, you know, a lot of our questions are scenario modeling. That what happens if I increase my investments in North America?
01:48:50
Speaker
What happens if I double up my increased investments in Southeast Asia? And all of this could actually be modeled on that. So you know the finance team should actually create a scenario modeler.
01:49:01
Speaker
And now you don't need to sanitize the data as much because you know these LLMs can take a bit of unstructured, unsanitized data. and structure it back out for you. but right So it first needs to get embellished everywhere in the organization.
01:49:15
Speaker
Then it needs to get embellished everywhere in the product. right That how does each text box, each default becomes a bit smarter. If you have uploaded me your input data, can I detect an anomaly that a table does not, it cannot be one centimeter by one centimeter by one centimeter.
01:49:34
Speaker
This is an input error. Otherwise it will create a loading issue. right So then the second is embellishment through each small functionality, each small screen throughout the product.
01:49:44
Speaker
right Because you see, fundamentally, you know each new tool is like electricity. If you're not going to put a bulb of an electric you know application at the end of it, you're going to shock your user. That's not what you can sell it for the sake of it. right It had to convert convert into application.
01:49:59
Speaker
And now comes the third piece. Where, so a lot of our users ah operate in multiple of their local languages, right? So fundamentally, if you think of it, you know, and in logistics, a warehouse manager knows that they want to look at all the orders that got delayed yesterday, correct?
01:50:21
Speaker
Every five years, we go and teach them it on a new user-friendly software.
01:50:28
Speaker
Our thesis is AI is the new UI. right That user interfaces will become obsolete into intent interfaces. So now what we have built is a full voice interaction with our system.
AI-Driven User Interfaces
01:50:42
Speaker
And when I say voice interaction, you can say stuff like, you know, change my drivers profile photos. And then it will find how what is the option in the desktop.
01:50:55
Speaker
And it moves the mouse. It shows you where to click, how to do that. And then it responds back to you. right So we are completely agentified using our system. Because what we realized is our users know what they they need to learn how to use it on the user.
01:51:11
Speaker
You know, it's much similar to like when initially computers came, you know, my father would come and say, hey, can you get me my SBI ballot? He knows what he wants to do. He just don't know how to use that user interface. But today AI can bridge that.
01:51:21
Speaker
And that's actually where we have, you know, done like the most insane amount of R&D this year. Right. And we're also launching it out as a separate company because effectively what we have built is a great computer use for embedded SaaS.
01:51:33
Speaker
Right. Now, any SaaS application can effectively launch a co-pilot in one day, not zero days, but one day. And this co-pilot will not be like actually just a question answer. No, it can actually take actions on behalf of you.
01:51:46
Speaker
Right. So we have like a wild version of it only for internal, you know. So while working out, actually say, hey, can you get me like a dal ghicary on swiggy? And I would do that. I would say, hey, can you get me a cab after the next meeting? I what would do that.
01:51:59
Speaker
like We're not launching a B2C extension publicly. That's only for our use. But what we're doing is launching as a B2B service. So, you know, because change management and training becomes such a complex thing.
01:52:10
Speaker
The idea of it also came via this only man. 12th February, we have ah we have actually a client in Mumbai, India gift portal. And they see you at the right concept that, hey, during 14th Feb, Valentine's Day rush,
01:52:24
Speaker
All of my external carriers will not be able to maintain the 30 minute SLA. There is this one carrier X which is typically expensive, but you override everything for 14-15 to this X. The CEO told their VP operations, their VP operations told my VP customer success.
01:52:43
Speaker
My VP customer success told my technical account manager. They did it on the same day. Everybody is very happy.
01:52:52
Speaker
When this technical account manager was doing it, i I was just so happy I was in the office and standing there and I realized they were doing it on the GUI. There was a GUI option to do this. ah the Graphic user interface. grey On our dashboard there was ah on our but on our customer dashboard, there was an option to just override it But it's hidden under the settings, it's an esoteric option. yeah know yeah And because you've made it so configurable, it will always be challenging to find everything. And now I'm like, we're going to spend effort automating these tickets. But the problem is this ticket should never be created.
01:53:27
Speaker
right? Tickets effectively become a more model of interaction in enterprise software. Yes, yes. And you want to just fundamentally append it, right? Your user interface, you know, so now everything effectively gets substituted with a search bar and a mic icon.
01:53:42
Speaker
right The future belongs to intent interfaces, not user interfaces. right We can now just explain our intent to the system and they'll get it. By the way, this is how traditionally every expert user has used it. right Developers use it over a command line.
01:53:57
Speaker
Gamers have a coac consee you know a console that drops down and we type in. Expert Photoshop uses type action script. It's just that for non-expert users, these were very unforgiving systems.
01:54:08
Speaker
With NLMs, we can create extremely forgiving text-based interfaces. right These are actually natural language interfaces, not keyboards and mouse. right So we've got to bring natural language back to computers.
01:54:21
Speaker
How would this learn everything which a SaaS product can do? Yes, so that's exactly I mean, ah so you know, it runs it scrolls all the interactions, it max of the site maps, right?
01:54:36
Speaker
It understands the intent, we have built a critique agent, which so for example, knowing when an action is done is a very complex problem to solve. Right. So doing the task is relatively easier versus knowing when the task is done, and if it is done correctly.
01:54:51
Speaker
right So we would use multiple LLMs, we cross-reference, we run critique. Then you know you have to have other tools. for example, you can say things like, hey, when this route plan is created, can you just take a screenshot and email me?
01:55:04
Speaker
Why should you be sitting in monitor? So it makes the whole system really fun. There's a lot of tooling infrastructure to make these you know things like screenshotting, email, and what happened.
01:55:15
Speaker
Then there's obviously, you know we rely on everything from Cloud to OpenAI for analyzing the DOM to understand. Then you've got to put in a lot of engineering to make sure nothing hallucinates. right Because now you're often doing actions which take 16 steps.
01:55:29
Speaker
You hallucinate at any one, everything goes off the table. right Your actions need to be explainable to the users. We don't want to call magic functions at the back end because we also want to educate the user.
01:55:40
Speaker
So yeah, it's it's ah you know we are launching it with some of our already existing customers. We are seeing like great, great adoption. We have launched a velvet rope for you know onboarding non-locus clients on this.
01:55:52
Speaker
And yeah, we'll be putting it up in q What you called Mycroft. what What do you mean? Velvet rope as in like, you know, we're not, we're not, because Mycroft, we made it initially for Locus and now we are externalizing it.
01:56:05
Speaker
So, and we have more demand than we can right now really address because we're a small team on that. So we are just, you know, you know how you have that velvet. It might be just a launch strategy. We are collecting interest and and each in Q1 going after selectively to put it out.
01:56:24
Speaker
So microsoft starting with locus microsoft Microsoft will start with the most expert user of Locus and then of her. I mean, this could be your AWS, like like how Amazon has the core e-commerce, but AWS is out of that.
01:56:40
Speaker
yes Yes. You can never take the AWS out of an AWS engineer. yeah So you intend to like build out a full team to sell this and to really scale this up as a separate standalone fast business.
Advice for New Founders
01:56:54
Speaker
so this is one of this is one of the ways, you know, Locus is using its capital. So now Locus is actually, effectively...
01:57:00
Speaker
started it out as a team, invested in this company, right? And then, yeah. so we have ah we have a great culture of both Acqui hiring good founders, as well as backing the team members who leave the organization, right?
01:57:16
Speaker
So we have like a bunch of other startups that have started out of ex-locus guys. And not only do we support we help them close their entire first round. We will put in a check, either personally or Wow. Amazing. Amazing. Amazing.
01:57:29
Speaker
So let me end with this. you know What's your advice to founders starting in 2024? I mean, i mean you know any advice any founder might have given two years ago also is not relevant today because of so many changes.
01:57:44
Speaker
So for a founder starting in 2024, what's your advice?
01:57:50
Speaker
It's the title of the book by Mark Anderson. The hard thing about hard things is that there is no pattern.
01:57:59
Speaker
Right. So don't unnecessarily search for one. Don't unnecessarily seek for one. Right. If you're starting one thing I would definitely say is get a co-founder. Right. Having a co-founder dramatically increases your probability of success.
01:58:13
Speaker
Apart from all functional needs, I would also say, you know, just just having somebody to grab a beer with when you're really, really down. Right. And that happens.
01:58:26
Speaker
So get a co-founder and there is no better time to build than today. Don't worry about patents, right? if If you were looking for a technical co-founder, you have AI. If you were looking for a marketing co-founder, you have AI, right? Like, so effectively, there are one thing I would say, and this is what I recently mentioned at the NVIDIA stage with Jensen as well, that all labor cost arbitrage is gone.
01:58:54
Speaker
So your only strategy of success has to be excellence.
01:59:00
Speaker
It can't be I will undercut the price or I'll be cheaper, right? Your strategy, you will be competing at a global level and your only strategy of success is excellence, right?
01:59:12
Speaker
But no better time than November, 2024 to Why is that?
01:59:17
Speaker
but why is that Because a lot of drudgery stuff has gone away, right? Like now when you think of an idea and you can conceptualize it and, you know, detail it, you can actually very quickly get to like, you know, a working prototype.
01:59:30
Speaker
Yeah. Right. So you even say, right, like, you know like if have if I've thought of a blog article, I would now put it up in a you know, 10 bullet points, focus on that, you know, over time, I've evolved my prompts that, you know, this the kind of style I write and all, I would give it and it would give me a very good draft back that I will then again go word by word, edit out a few lines, right? So That whole inertia of just sometimes sitting and you know nothing coming out that has gone. So your response cycles have become much shorter.
01:59:59
Speaker
So everything is a lot more fun today than it was a year back, right? Everything from writing a hiring post to writing a code, to writing an organizational memo, right? To learning about anything.
02:00:11
Speaker
Right. Like I use like that advanced voice mode quite a bit, right. by While I want to move to learn about something. It's yeah. So everything is ah just a tiny bit more fun because it's a bit more convenient. So that's what i'm saying.
02:00:22
Speaker
Like no better time than now to build. ah What is your personal content consumption habit? How do you consume content to learn stuff? I, um so I have a dedicated time called wandering.
02:00:37
Speaker
Right. So i I split my time and like, you know, workout, work, family and wandering. And I really cherish this, you know, time and wandering.
02:00:49
Speaker
One of the things, one of the best things I did over over the last three years for my personal content and mental health was to gravitate largely towards long reads. Right. So often now I will just send stuff to Kindle.
02:01:02
Speaker
and read it at night on a Kindle, you know, outside of other screens. Right. oh Yeah. So I do, I do want, yeah, I think this, that wandering is a very critical part of my day and something I truly relish.
02:01:17
Speaker
ah So wandering is like, you just read anything that interests you. and So I'm typically, you know, always have topics. I'm, yeah you know, currently oh anxious about not knowing enough. For example, right now, one of my things is, you know, the human computer interaction, right? Like the brain machine interface. This was something I used to have a lot of interest during my undergrad years, but now I haven't really caught up with all the stuff that Neuralink does.
02:01:44
Speaker
So been reading all of those bio papers. Yeah. Wow. Fascinating. I noticed you you don't have anything there for entertainment, like for your Netflix time and all.
02:01:57
Speaker
um My entertainment involves being outside, right? So I love going for a run with my dog. I do a lot of gardening. This year, i was able to you know produce 20% of my vegetables during the summers.
02:02:10
Speaker
Wow. I, as an amateur, I race tracks, I fly a single engine turboprop. Yeah. So all my fun involves being outside in nature. Wow. Amazing.
02:02:20
Speaker
Inspiring. Thank you so much for your time, Nishit. It was a real pleasure. Same, Akshay. Thank you. It was a wonderful conversation and I clearly didn't realize where the two hours went.
02:02:31
Speaker
And thank you for holding up that energy on the table.