Transcript
Speaker: You're listening to From the Horse's Mouth, intrepid conversations with Phil First.
Speaker: Welcome to the very latest edition from the Horses Now podcast. I'm your host, Phil First, and joining me today is um somebody I think I first met back in around 2002, I think it was.
Speaker: um And we managed to stay in touch as ah kind of friends and peers over the course of the last couple of decades. So it's great to have him um finally join this podcast. And Ritesh Adnani, you've had a unique career across both services and software.
Speaker: you know So looking back, what has been the most important lessons you've learned along the way until your current role running First Source? and Maybe you talk a little bit about maybe who are the people who influenced you the most and some of the defining moments ah that have really shaped your career today.
Speaker: ah Thanks for having me, Phil. And sometimes knowing how long we've known each other is is almost a giveaway for how old we are. But no hope hopefully we've seen the trials and tribulations of the industry itself, right?
Speaker: and When you win your pitches, predate the iPhone, you know.
Speaker: Yeah. No, but you know, it's ah it's been an interesting journey. i I'd like to believe that I've had a career that almost has swung between two poles, if you will, right? On one hand, running the operational engine of...
Speaker: large services businesses, and on the other end, trying to build something new from a blank page itself, right? So they've oscillated between both these extremes in a way. um and And if you go back career-wise, you know, ah my formative years in Citigroup and Infosys were really the foundation to where I find myself today. And ah particularly if you go back in time to Infosys itself, right, growing that business from what was a $100 million dollar business, having a ah
Speaker: front front seat, if you will, to seeing the growth from then to an $8 billion dollars business was quite ah quite a journey, including seeing the early years of what BPO would become.
Speaker: ah But if you look at the last decade, that decade that followed was almost... ah different in a way because I spent my time with with a firm called ISG and went from there to Seamless Health, went from there to Tech Mahindra and then Unifor. But in a way, if you if you now go back in time and abstract it out, each was a different bet on a certain layer of the services stack itself, right? Whether it was vertical software, whether it was healthcare care services, whether it was design and digital, or whether it was conversational AI. And and now when I look back, um
Speaker: you know, from a journey perspective, what it's really taught me is ah three things, right? Number one, ah just being very comfortable with ambiguity. And I go back in time to probably one of my first first formative experiences, you know, when I was... ah sent to Citibank and Calcutta to manage their foreign exchange operations you know at the age of at the tender age of 23. This was an operation I'd never run.
Speaker: It was in a unionized environment, and I was replacing somebody twice my age. ah I had to learn the darn thing from scratch. And I think what that teaches you an enormous comfort with ambiguity, which eventually I recognized was a superpower, if you will, for me, from my vantage point. I think the second thing that ah that I would say is um speed beats perfection.
Speaker: um And, you know, my colleagues at FirstSource will testify to that in some form or fashion where, ah you know, i keep telling them that speed is the enemy of perfection. Get me something half-baked and that that will go a longer mile than ah waiting forever to get the perfect story being told. I'd rather get something half-baked than iterate. And I think that... that lesson probably holds true even more so in today's day and age, ah given given that speed is a competitive moat, if you will. And third, I think, is be a partner of relevance, not necessarily a partner of scale. I mean, clients today don't necessarily need you to be the biggest player in town, but what they want you to be is the most relevant for the problem that they're looking to solve. And and I think if you're able to continuously bring these three to the table, I've seen that they they hold you in good stead.
Speaker: It's interesting talking about, um, you don't have to have everything fully baked before moving, moving forward. Right. And, um, you know, obviously today's feels like an inflection point. You know, some people say it feels like 99 over a again. Um,
Speaker: with what we've seen happening with AI and its impact on people-driven services and people and scale and all these sorts of things. So how do you see the future of the BPO industry now evolving over the next five years? You know, let's let's be realistic about this. You know, what's going to separate the ones who succeed versus the ones who are going to just really struggle to adapt?
Speaker: ah That's a great question. You know I think BPO is heading into, in in my mind, the most consequential five years of its history itself. ah The labor arbitrage era, which it's really defined and built this industry, um I think it's passed its sell-by-data as a basis of competition.
Speaker: ah and And I know, you Phil, you you framed this ah correctly as the services as a software era itself. And I think that framing is is absolutely correct, right? What I would add, though, is that the winners and losers here will not be separated by whether they adopted AI or not. Everybody's adopting AI. But I think they will be separated by whether ah they've had the courage to cannibalize their own model before somebody else did it for them.
Speaker: um And I think that a... three things that I think will separate winners in this era, right? um First, the willingness to redesign the workflows, not just bolt AI on top of what might be a broken process.
Speaker: You know, when you rebuild a workflow rather than bolt AI on top of it, the gains aren't ah incremental. They're truly transformational. And I think that matters. um I think the second one is taking accountability for the outcome, right? ah In my mind, the shift from a,
Speaker: labor-based delivery to a nonlinear outcomes-linked delivery ah is quite fundamental because the unit of value stops being the seat and starts being the result.
Speaker: That's a fundamental rewiring of how how this industry is going to be. And and I think the third is ah the willingness to be vertically deep rather than horizontally broad.
Speaker: um An inch-wide, mile-deep will consistently beat a mile wide and an inch deep. And we see that in our business, right? I mean, ah we do well where we have an end-to-end understanding of the processes in the markets in which we operate ah because that's where we can we can underwrite the outcomes itself, right? And and I think conversely, therefore, the firms that will struggle in my mind, will be the ones who are still measuring themselves based on either the seats added, the headcount added, still pricing inputs, still running pilots without a P&L accountability.
Speaker: And quite honestly, I do believe that the market will either absorb or pass them by. There is no soft landing for a labor-first model in the current environment that we are in. And I think that's going to be the harsh reality that I think we're all going to encounter.
Speaker: Yeah, I mean, it's interesting. I had our advisory board meeting yesterday and and some folks on there you'll know well, like Frank D'Souza, Tiger, people like that. um And we were sort of talking around the deflationary impact on share prices on services firms. And you know, there was a common common view that there's an absence of the CEOs and the public services firms acknowledging AI's deflationary impact and um not having a strong narrative um that could that would boost stock prices. And this is why there's this sort of gap between kind of Wall Street and Main Street. And it did make me think when we were talking to you, was around, you know, you talked a bit about un-BPO and
Speaker: First source was a pioneer in introducing this to market and all this sort of stuff. So what's your view on this whole, are we not just communicating this well? um What are you guys trying to do? And and and um how do you see services getting out of this sort of hole that we find ourselves in at the moment?
Speaker: so So that's a lot of different you know questions, but let me try and unpack that into a few, right? So first is when we ah launched on BPO in February of 2025, we were very clear to call it out that the labor arbitrage era is past its sell-by-date. This is no longer the era of faster, cheaper, better. And and the way this industry has grown over the last 25 years, you know where scale, standardization, and all the elements that that came with it were the defining attributes of success itself. And and I think OpenAI did that when they launched ChatGPT in November 2022, that the first winds of change were ah necessarily going to play about, right? and
Speaker: And therefore, recognizing and acknowledging firstly that labor arbitrage is the basis of value creation, headcount growth or FTE pricing as the unit of growth are no longer relevant metrics, right? if you're in a state of denial around that, I think that itself is is a challenge. and And I'll come back to the the point around, you know, how do you therefore ah manage or interact with investors given given that this context is significant because the last 25 years, the industry has operated in this manner itself, right? I mean, i I still know when I go and meet with institutional investors, they're sitting out there with a model and they're saying, have you guys added account because they're looking to plug it in a
Speaker: ah in an Excel spreadsheet and then use that as a lead indicator of revenue growth itself. And that that model's got to shift, right? But I think one of the things when we launched on BPO was not to say that the industry is dead. It was around saying that the traditional ways of working of the industry are over and the the traditional model is outdated and you've got to reinvent with the times. And the the fundamental question that we were looking to ask there is that what do you do when the ah when then this industry...
Speaker: ah pivots around intelligence and not labor as the primary unit of value creation. And I think that's a fundamental shift that that I think we've all got to recognize. And I think over a period in time, people are starting to smell the coffee, wake up and smell the coffee and say that, hey, that is that is the case. Now, one of the things that we ended up doing along with this is also saying that ah un-VPO is not enough. So we launched what we called as intelligence that operates. And the thesis behind that was to say that we will be the full stack partner. We will design the operating model. We will go out there and build the intelligence layer. We will ah run this in production, but we'll also put our name on the outcome and we'll underwrite the same.
Speaker: and And the idea that, I mean, I was very clear in this, if we can't underwrite the result, then we can't put our name behind the same. um and And then when we launched Kairos, the idea behind that was to say that it's an embedded agentic operating system.
Speaker: ah It's not a layer that's added just on existing operations, but it's integrated with them. and And this was critical because in a way, it was intended to say that it's model agnostic. ah any Any model can work with it.
Speaker: Number two, it doesn't matter. ah you're You're not looking at seats, but outcomes. And third, and this is the part that I think you've pushed ah everybody. and I know I've seen some of your previous podcasts. You've pushed every CEO on on is, you know, ah what does it mean in your environment? Have you run it on yourselves? And we we ran it on ourselves before we brought it to market, right? Because if it didn't perform well in our own operations, we didn't have a stake ah in the ground to claim and run it in anybody else's, right? And we've seen some pretty astounding results with with the same. um you know We ran this for a UK fintech where you know we've we've seen more than 80% plus faster KYC boarding. We saw this in the health system where denials rates have ah have gotten remarkably better. And these are critical, right? Because the key thing to recognize in this environment is that AI agents are almost like a surgeon.
Speaker: a surgeon is only as good as the operating room around them. And I think with Kairos and with intelligence that operates, where we wanted to bring ourselves is to say that FirstSource will bring the operating room to you.
Speaker: ah The AI agent layer will be commoditized over a period in time, and that's not where the differentiation is going to happen. But going going to the other point that you made about the ah about the investor side itself,
Speaker: Look, I know you you called it around the, I read this LinkedIn post about what you said around the J-curve itself. And I think that's exactly right. I would only add one wrinkle to it, right? Inside our business, we're not necessarily seeing what I would call as a clean J-curve, but what I would call almost as a K-shaped growth dynamic. Let me, ah you know, unpack that in terms of what it means. On one hand, the traditional steady state business will continue to keep declining as clients shift away from labor-based pricing to more nonlinear commercial models. But on the other hand, you're going to sign up new transformative deals. But those deals don't necessarily follow a linear ah ramp curve itself, if you will.
Speaker: They're transformative. They ramp in phases. um Unlike traditional steady-state work, So while they strengthen long-term revenue visibility and durability, revenue conversion is typically spread over a longer period. Now, that's an important one to bear in mind and explain to investors because it's extremely critical to change what the leading indicators are. Headcount growth is no longer a proxy for growth.
Speaker: And we you know we started talking about this in our previous earnings call. We said ah two new metrics are important. The percentage of revenue you get from nonlinear constructs and number two is the revenue per employee. Just as an example, our revenue per headcount, revenue per employee is up about 12% over two years. And we expect that to widen actually going forward because we are tracking that like a hawk.
Speaker: Number two is be honest about the staggered ramp with investors. the revenue is not necessarily gone. It might get deferred over one or two quarters or thereafter. ah What investors will tolerate is a reshaped revenue curve. What they will not tolerate is surprises.
Speaker: I think the hardest part year is internal, more than anything else. Telling your sales force, your solution leaders, your operating leaders that the old commercial constructs are over.
Speaker: The cultural change is the real J curve and it precedes the financial one. right ah I think if you're able to consistently do that, you know ah the investor community gets around you.
Speaker: What they don't like is surprises because surprises kill multiples. It's kind of like we talk about services as a software. And it's really this whole dynamic of how do you scale a solution, which is what investors care about.
Speaker: And that feels like what you've said is right is cultural, is that we're selling ah a full stack. We're selling a platform. And it's like, um i I remember SAP back in the days of growth, particularly where no one seemed to realize that SAP itself had like a $7 billion dollars consulting business to help clients benefit from SAP.
Speaker: But the whole business was centered around selling a platform. Why can't we make that shift in the services industry towards selling like a stack or a platform? And then the mindset needs to be we're servicing this platform to our clients and we need to price accordingly around the platform.
Speaker: We need to price around the value, not around the people anymore. Is that what you're saying? When you think about the cultural shift, we have to start thinking about selling the stack, not selling the people.
Speaker: Yeah, I think it is. And, and you know, it'll it will go back to something which ah which I've been fortunate through the course of my professional career to be both on the software and the services side, right? And I think, in a way, when you look at how these these two ends of the spectrum operate, they're fundamentally different. and And I think both software and services firms can learn a lot more from each other than otherwise, right? So, I mean, if I were to go back and reflect on how how these industries have built themselves and where services can potentially take a leaf out of, you know you get the example of SAP and the consulting business that they that they had built and so on and so forth. But if you take services businesses and and and say, what can they take a leaf out of traditional software firms, right? First is a product mindset.
Speaker: Services businesses haven't been built with a product mindset, right? Software firms intentionally build ah version 1.0, which is not complete. And then they iterate. right ah services firm try to ship 1.0 almost perfectly as a statement of work.
Speaker: i mean, that's a fundamental difference in terms of how how these two archetypes have evolved over a period in time. The second is software firms think about this almost as a compounding architecture right because they design it in such a way that every customer makes the next customer cheaper.
Speaker: That's how the SaaS business evolved. The SaaS business was all about why were investors giving a high premium ah from an ARR standpoint? Because they said, you know you keep adding customers and over a period of time, the incremental cost of adding that is is is is about the marginal cost of adding that keeps coming down. right Services firms, conversely, designed every engagement almost as ah on a bespoke fashion, where the second and engagement costs almost as much as the first one.
Speaker: but And that's a fundamental disconnect in terms of how how these have ah have organized. right And one of the things we're trying to do when we're talking about compounding outcomes is we are deliberately and intentionally re-architecting first doors around the compounding advantage.
Speaker: Because you're saying that is a big shift in terms of what needs to happen. And third, I think this is an ah opportunity for both ends of the spectrum. right The um services firms have traditionally...
Speaker: build on a time and material basis. 75% of this industry is still time and materials, and that's a problem, right? But conversely, you also have software firms which are subscription-led, ARR-led, right? And there is a vast middle which is which is sitting, which is around the nonlinear side that services has barely begun to inhabit, right? So I think that's the challenge that's there. But I think there's also a reverse problem, right?
Speaker: Which I think is important for ah I mean, and you asked the question around investors, I think, you know, one of the other things that services firms need to do a better job to explain to investors also is that two things are critical. Domain debt, number one, which software firms don't necessarily have.
Speaker: And software firms oftentimes underestimate how much of an AL deployment fails, not because the model is wrong, but because the workflow it sits inside is misunderstood.
Speaker: Service for firms have decades of muscle memory. and And, you know, if you have a lot of domain depth like we do on the vertical side, I think you know how work actually gets done. But the second point that that is there is also the change management. The biggest thing to address here is the change management associated within the workforce, right? You can have an AI model that is technically sound and culturally rejected.
Speaker: And that will still be a failed AI model. what What I think my crossover experience across these c two industries has really taught in summary right is software taught me that the second customer should cost half of what the first one.
Speaker: Services taught me has taught me that no model survives a workflow that it doesn't understand. I think the firm of the future is one that speaks both languages. Do you think... because of this Because of this convergence, um I mean, we've seen the prices, the valuations of services firms, especially those on the U.S. market, absolutely plummet.
Speaker: Do you think we're now going to see a the first real consolidation between different firms and services firms? like you like What's stopping... I don't know, SpaceX has spent $60 billion on Cursor, and you can now buy Accenture a lot more, or you could buy a cap Gemini or Cognizant, these companies. Do you think we might start to see some real shifts in how these companies are made up because the dynamics are just just completely changing beyond recognition?
Speaker: I think there are going to be some interesting crossovers for sure. that i would I would fully anticipate that. But I also think that people are going to be mindful of ah of how they do this, right? ah I think in this environment, we are firmly in the throes of of the innovators dilemma. And if you go look at the innovators dilemma and what Clayton Christensen said ah out there,
Speaker: It traditionally tends to impact larger incumbents more. I mean, history has shown that over a period in time because they are slower to adapt and respond and so on and so forth. And it it spawns the way for a newer set of challengers challenger brands and startups to emerge who can disrupt because they can play on offense.
Speaker: um ah I think that will potentially come in the way of what kind of crossover consolidation could end up happening. Because, you know if you're acquiring a firm which has got 400,000, 500,000 people, ah you know what? Changing the wheels of ah of an organization of that size and scale is not going to be easy.
Speaker: And therefore, are you saddled with more... process debt and people debt and talent debt that you need to deal with versus saying, you know what, I could build this ground up itself and maybe it might not be that hard. So I think that's the dilemma and the trade-off that I think firms are going to look at in terms of how they acquire. There is going to be probably some um acquisition on the context of revenue growth too, because, you know, if revenue is going to slow down for for folks, they might say use acquisition as a tool to try and, ah you know, short the top line up.
Speaker: while they continue to make the changes and in in the way they operate itself. So you probably see some, some some ugly but I think it's a combination of these kinds of things that, that will, that will emerge um across the board. I mean, SaaS companies will probably acquire some AI native startups. You've seen that with, with the different players in the marketplace. You saw that most recently with what so but Salesforce did when they acquired Fin, um,
Speaker: ah You know, you're you're going to see more and more of, you know, interesting crossover plays that could emerge. But, ah ah you know, ah that's going to be the opportunity. So four or five years out out from now, this industry is not going to look very similar to how it does today. I think there's going to be a lot of change.
Speaker: And and you're you're putting more thoughts in my head as we go through this compens conversation. But um yeah we put out a lot of that research on the the cost of enterprise enterprise debts, like talent debt, people debt, data debt, tech debt.
Speaker: The one thing I think we're discounting here, which you've really highlighted, is a lot of this debt is sitting within service providers. ah If service providers have 500,000 people or whatever um delivering certain number of clients, they're trying to hit a certain margin, and they're struggling to scale that business without constantly adding more people, then they've got their own debt, which they haven't figured out how to know deliver a more standardized capability without having to constantly increase labor. So it almost feels like there's a huge debt debt problem on the services side, as well as the client side, and both sides need to fix these issues if if we're ultimately going to come out into a more streamlined environment, right?
Speaker: You know, and I don't know think it's so important to understand in the context of how ah how automation itself has evolved over over a period in time, right? Because ah what we're experiencing now is, I think, fundamentally different, right? So let's take go back in time to how this industry has evolved and why this time is different, right? Right.
Speaker: The last 20, 25 years, um you've seen different waves of automation, right? We started off doing some stuff on the point solution side. There was workflow automation. RPA was the, I remember, i mean, you were at the forefront of coining RPA and getting it you know recognized out there. But you know the the different waves all gave you some incremental benefits. But what they did is they did stuff faster, cheaper.
Speaker: But the work of the the nature of work was of the same shape. and and and I want to put that same shape almost in quotes, right? I think what we're experiencing now with generative AI and and thereafter with agent tech AI is the first wave where the it changes the shape of the work itself.
Speaker: That's a fundamental difference because now you're in an environment where two or three things are happening, right? Number one, Agents do not just assist work, they can execute it. You know, I said this on our last earnings call, right?
Speaker: Processing transactions from process throughput to decision quality. This is no longer about you know us getting paid based on the number of transactions we process or the number of butts on seats that we have, but around whether we did the right took the right decision each time for a transaction. That's a fundamental shift in terms of how one should think about it, right?
Speaker: And third is the fact that learning is compounding. I mean, I referenced it earlier also, right? When we do work for the first time, we gather information about it. then becomes a compounding advantage when we do the same process for a second client and a third client and and a fourth client. That's a fundamentally different economics, right?
Speaker: In the labor model, every new client started the cost curve from scratch. In the agentic model, every new client benefits from the previous one. There are fundamentally different ways of how one should think about it, right? And and therefore,
Speaker: The second client in the labor model costs you almost as much as the first one. In the agentic model, the fourth client is where the compounding value and the benefits, the unit economics actually start. How do you build a business recognizing that when you have headcount which the service provider is saddled with? I think that's the $800 billion pound question for for organizations today in terms of how do you deal with this?
Speaker: This takes us into this whole debate around um yeah AI eliminating jobs or elevating them. um You've had a few years experience doing this, and I know you're personally very yeah very very involved in these tools. But um what do you think the workforce of a leading services company is going to look like in five to ten years' time based on everything that you've seen in the last couple of years?
Speaker: There's a lot of lot of talk around AI eliminating jobs. There's also a lot of innovation around AI elevating jobs. and And the honest answer is actually both both these things will happen. Some jobs will disappear.
Speaker: um Some jobs will get elevated. ah And new jobs that don't yet have names will get created. That's just the nature of how this is likely to evolve, right? ah What I would start off by saying that anyone who says that in this industry that AI is only additive with selling snake oil.
Speaker: And anybody who says that AI is only subtractive doesn't understand the demand side of the equation. I think both both these things can kind of coexist in a way, right? ah
Speaker: Given that context, right? ah Five or 10 years out, in my mind, there will be three things that will probably, three or four things that will be probably true about the workforce of a services company, right?
Speaker: Number one, fewer people but doing higher value work per unit of revenue. I'll repeat that again. Fewer people doing higher value work per unit of revenue. The reason why that becomes critical is the pyramid will flatten into a diamond.
Speaker: The narrow base of repetitive tasks will get replaced by agents. The expanded middle is where the new jobs lie. And these new jobs could be people who are workflow architects, people who are AI trainers, people who are decision auditors, people who are context and policy engineers.
Speaker: Some of these are just starting to come emerge from the woodwork now, right? But this entire notion of fewer people doing higher value work per unit of revenue will become accepted, ah certainly for the successful services companies of the future.
Speaker: I think the second part of it is domain depth will become the single most valued credential. and And I want to elaborate on that, right?
Speaker: Generic process knowledge is going to be table stakes. the The AI model spits it out today. I mean, you go and ask your favorite frontier lab and you want to understand a workflow process. It's going to give you the basic stuff.
Speaker: AI has that today. But knowing how a Medicare claim moves through a payer, That is a scarce skill. So I think that's where there is the rubber meets the road and the domain depth becomes the single most valued credential. And I think the third thing that I think is going to be a different ah differentiator for um successful companies as they, you know, or the workforce of a services company is accountability will replace ah activity as the measurement of contribution.
Speaker: You know, today I'm, you know already asking our employees, you know, every single day, What value did you add to your customer today? Right? ah that This is not about what activities are you supporting. If you're in a contact center, it's not about, you know, how many calls did you take? What's your average handle time?
Speaker: I mean, those those metrics are relevant. They're activities. It's not to take away from that, but they're input metrics. They belong to the agents. I would say the value metric belongs to the human.
Speaker: And I think that if you start thinking about the the workforce of a services company is five to 10 years out, these three things have to be critical. Fewer people doing ah higher value work per unit of revenue. Number two, domain depth. And number three is the ability to think about accountability, replacing activity as a measurement of contribution. These are fundamental shifts. This is not the way this industry has grown over the last 25 years.
Speaker: So how do we make this change? Because it's easy enough to talk about This is what, the you know, the middle is going to become the new the new orchestration layer.
Speaker: We're going to change how we operate. The middle management is not going to be like a relay of information up and down the stack. um So surely then this moves to leadership, right, in terms of... um How has AI changed? Like firstly, one, how are you personally operate as a CEO?
Speaker: And how does this change how you work with your leaders within your company so you can make this evolution that you're talking about? So a few changes in terms of of how I operate itself, right? First is, you know, I do run my thinking ah through AI now daily, you know, whether it's ah preparing for an investor call, whether it's reviewing the structure of an acquisition target, whether it's stress testing a board narrative.
Speaker: ah I use AI as a first pass collaborator, not for the answer, ah but for the better question. And I think that's the important differentiation that I want to want to call out, right? the The folks who are using AI to generate output are missing the bigger picture.
Speaker: AI's gift to a CEO is provocation, not production. Use it to ask you the question you didn't think to ask.
Speaker: And I think that's that's one fundamental ah thing in terms of at least how I've been i've been reflecting on this point itself. But the second thing that I think is critical is ah I've shortened my own decision cycles almost intentionally itself, if you will, right? AI has compressed the time it takes from question to draft.
Speaker: And oftentimes what it does is the temptation is to fill the time it gives you on your calendar back for more meetings. I've almost done the reverse counterintuitively to say, can I extend my thinking time?
Speaker: right? The CEO of a services firm in 2026 cannot afford to make the ah decision the same way that he or she did in 2022, right? And third, I think is I've made AI fluency almost a non-negotiable for my leadership team itself, right? You cannot lead an AI-first organization if your own AI literacy is below the median of the room.
Speaker: That's a problem straight off the bat. Because, you know, technology, if you're challenged by the technology yourself, right? And The reason why it's important is because the team can tell. They know, you know, if somebody is not getting it, right?
Speaker: Therefore, my advice to our leaders is first and foremost, use it daily. Use it on your own calendar. Use it for your meeting. Use it on your writing. Use it on your decisions. But second, also be very honest about what you don't know and learn in public with your team.
Speaker: And that's perfectly fine. You know, ah we have a bunch of, ah you know, some of these young kids who've joined us, computer science grads. I spend time with them every week. And say, what are you guys working on? What else should we we be thinking about? And sometimes it's just a question of getting a osmossis ah reverse osmosis from them when you know you just learn from some of these kids and they actually give you a heck heck of a lot of energy by just the various things that they are ah trying out and tinkering with and so on and so forth.
Speaker: And I think the third thing that that I would say is, commit I mean, this sounds cliched, but committing to lifelong learning in this environment has become even more critical. Because the shelf life of certainty has never been shorter.
Speaker: You know, so I think from that vantage point, we we don't really have too many too many alternatives. So, you know, if a leader is not doing these three things on a consistent basis, I think they will they will struggle.
Speaker: Yeah. Yeah, I know. you're You're getting me thinking a lot. But it's also fun. and These tools can be very frustrating, but they can be amazing at the same time. I mean, I can go to an analyst and say, yeah oh, did you get Claude to write that with you?
Speaker: And they'd be like, oh, my God, no, no, no. It's all me, honestly. I went to one of my sales guys the other day, and I said, great proposal. It reads very much like Claude. And he goes, yeah, it's really good, isn't it?
Speaker: And it's it's a total mindset thing because he's just thinking, I want to get this deal closed and this thing can help me get a proposal written. That's no secret, right? Whereas people who are like trying to use this as a knowledge tool and I'd rather, you know, that consultant or analyst turn around and said,
Speaker: Yeah, how do you think I should use this tool better so I sound more authentic or something like that? But I think it's a voyage of discovery as an organization. i mean, I like talking to my CFO as well. I'm like saying, you should really use this stuff in Excel. It's brilliant. like i said I said, this is revolutionary as my PowerPoint experience. I've been waiting 30 years to have something like this in PowerPoint, right?
Speaker: Absolutely. Couldn't agree more, yeah. This has been a really honest conversation. I think a lot of people appreciate some of the views you've shared. and And i did mean I did mean what I said earlier about i think CEOs in this industry just need to be more open about, hey, look, we're under siege. This model doesn't work anymore.
Speaker: We need to really change things up. And I'm not hearing enough of that. And I think hearing from you and just talk about un-BPO and things, you know, I think it's bold. I think you're trying to sort of say to everybody, look, we're trying to get away from that model because we're trying to be something else.
Speaker: So being realistic, when we sit down in 2030, isn't that far away now um what do you think would need to happen for you to say the service industry successfully reinvented itself and we didn't just see a lot of these companies crash into insignificance you know um what do you think um would really need to happen and let's say in the next sort 12 to 18 months but we talked a little bit about lead indicators and and and what what wouldn't start needing to change in terms of the narrative that, you know, CEOs are taking to ah boards and investors and and employees in terms of how how this is going to evolve. But I think if you're sitting down in 2030 and we say the services industry has successfully reinvented itself, I think three things would be would be probably true, right? Number one,
Speaker: um a nonlinear commercial model would be the most dominant model. So think about that today. 75% of this industry is time and material. The fact that that will start shifting is going to be critical. That needs to flip.
Speaker: The majority of ah revenue should come from a nonlinear model, nonlinear construct. Without that, the labor-first economics will keep us trapped. I think that's one one key metric that I would i would hang my head on.
Speaker: I think the second one is that the revenue per employee curve would have visibly decoupled from the headcount curve. across the top firms visibly, audibly. It's there in the filings. And that would be a verifiable signal that the industry has moved.
Speaker: And I think the third one that I would put is services firms would be where the most ambitious AI engineers, and and this may be a you know a reach at this, may seem like a reach at this point in time, but the most ambitious AI engineers would say, I want to work out here. And here's why, right? Because it's not as a fallback to a frontier lab, but it's as a first choice because you know what? The most interesting AI problems solving for a messy real life workflow,
Speaker: putting that at scale in a production account environment with accountability is pretty darn exciting. And if you're able to do that, that that would be success. But you know the flip side of it is if in 2030, we are still pricing seats, still calling ourselves BPO, is still selling AI as a feature on a labor contract, still measuring success and heads added, I think those are all indicators that the industry hasn't really moved much.
Speaker: I liked what you said about um AI problem solving. Maybe that's the next phase. Go from services to problem solving because that's a lot. A lot of what we're all about right now is fixing a lot of what we do wrong to so we can do it differently. You know, I'll just say one thing there, right? you Look, the technology will arrive on schedule.
Speaker: The courage to use it will not. And that's the binary that I see for 2030. If you put a magic wand out there five years down the line. Well, the courage to use it is a perfect way to end this conversation. And I enjoyed it very much, Ritesh.
Speaker: I think you're going to be with us in London in a few short months as well. I hope you get chance to meet our London crowd before long. And um it's great having you today.
Speaker: but Thank you for having me as always, Phil. Good to see you. Good for catching up soon. Thank you for listening to From the Horse's Mouth. Don't forget to subscribe and like wherever you listen to podcasts.
Speaker: Got something to add to the discussion? Drop us a line at fromthehorsesmouth at hfsresearch.com or connect with Phil on LinkedIn.
