Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
Why Big Contractors Are Building AI Tools In-House image

Why Big Contractors Are Building AI Tools In-House

The Off Site Podcast
Avatar
0 Playsin 3 hours

Join Jason and Carlos in Episode 94 of The Offsite Podcast as they explore three major developments shaping the construction industry:

๐Ÿ›ฃ๏ธ Abidjan-Lagos Corridor: They dive into the $15 billion, 1,000-kilometre coastal highway connecting five West African nations. The duo explores funding challenges and potential contractors for this transformative infrastructure project serving 300 million people.

๐Ÿค– Contractor AI-Powered Tools: They explore how contractors like Skanska and Granite Construction are using AI to build custom solutions like "Asphalt Guru" and "Safety Sidekick." Does this fundamentally change the build-versus-buy equation for construction technology?

๐Ÿ’ฐ Outcomes as a Service in Construction: They analyse the emerging pricing model where companies pay for results rather than software access. The hosts examine the benefits and complexities of this approach, from risk allocation to outcome measurement challenges.

Key Timestamps:

00:00 - Introduction

01:42 - West African Highway Project

07:58 - AI Development in Construction Companies

24:23 - Outcomes-as-a-Service Deep Dive

Check out the Off Site newsletter here

Follow Carlos on Linkedin | Follow Jason on Linkedin | Check out Aphex

Recommended
Transcript

Introduction and Episode Context

00:00:00
Speaker
We know from like construction ourselves, people love to build like macro spreadsheets. So this is like that on the heat. So you can see there's gonna a massive cohort of people really excited to go and build everything. Did you all just say that on heat, by the way? just I'll just pause it. I've never heard. I would have said like that on steroids or that on heat. Same thing.
00:00:25
Speaker
Welcome back to the Offsite Podcast. I'm Jason Lansini, joined once again by Carlos Cavallo for episode 94, getting close to the the big 100 of the Offsite. Carlos, both of us in a ah kind of warm London, both of us coming off maybe a few drinks last night. how are you feeling, mate? Since having a a kid, midweek drinking gets exponentially more difficult. So, yeah, we had a we had a staff social on Wednesday. And then we had, I went on a social with a Balfour Beach team last night. So, um yeah, a solid end to the week. You might need to bleep that because you might have got some of the team in trouble. So we have to bleep out the contract.
00:01:11
Speaker
So good end to the week. But it's nice to be able to go out and have a drink in the sun for sure.

West African Infrastructure Projects

00:01:15
Speaker
yeah the So today what we're talking about is ah we've got $15 billion mega project in West Africa. um We are talking about whether the contractors are turning into AI development houses.
00:01:30
Speaker
Spoiler alert, probably not. And can construction tech or should construction technology pivot to this new foundation? Flavor of the Month Outcomes as a Service. Should we just dive in?
00:01:42
Speaker
Let's do Let's do it. Got it. So ah for background, the offsite newsletter that was published from the team last week um dives into this massive a highway project in West Africa.
00:01:57
Speaker
which is probably one of the most ambitious infrastructure projects on the continent. Uh, it's about a thousand kilometers long highway that connects five West African nations. Um, and it's designed as a six lane coastal corridor,
00:02:12
Speaker
ah that's supposed to link something like 40 million people across um some of the most economically dynamic cities. We have talked in the past about um projects that sit across many countries or ah or regions and the difficulty in coordinating and delivering those projects. Carlos, it seems as though the same pattern repeats here, but it's a very ambitious and impactful project in Europe probably one of the you know one of the biggest growth regions over the next decade two decades three decades
00:02:48
Speaker
Yeah, absolutely. i think, um yeah, I guess for ah for a bit more context, anyone slightly geographically challenged, it's ah it a long stretch of coastline on the sort of southern coast of Western Africa.
00:03:01
Speaker
um Current travel time is something like 15 hours to do that thousand kilometer on existing roads. And that could be cut down to as short as eight. So it's a significant our reduction in travel time by having like a free flowing major highway.
00:03:18
Speaker
if and it's Sorry, just if i if I do quick dumb mathkin napkin math, it's about a thousand kilometers. so so You could do like 120k. Highways have a max speed in some of those countries, at least in Nigeria, of 100 kilometers an hour.
00:03:32
Speaker
So 10 hours. Significant. Did we just come across a classic example of government infrastructure projects over-promising on benefits? Yeah. Yeah, I think so. Like, yeah, we're going to go down to four hours, but don't worry about all the regulations that mean you can't actually do that.
00:03:50
Speaker
Yeah, there's ah there's a huge emphasis on trade with this. So they've got like dedicated lanes for haulage, freight, things like that that are moving. So it's like a main road for not just the local population, which I think you said was 50, 60 million, but the wider economic area is something like 300 million people that could utilizing that space.
00:04:11
Speaker
And so yeah, from a growth point of view, i had listening to a podcast the other day and it's something like one in five babies will be born within this region by like 2050 because of the the growth rate of these

Highway Project Design and Economic Impact

00:04:23
Speaker
these countries. So um you can imagine, yeah, the need for infrastructure with that sort of growth rate is really needed.
00:04:29
Speaker
So what's the story on the funding on the project? Yeah, so there's a ah bunch of sources in terms of um where the money's coming from. I think it's $16 billion dollars is the estimated um cost.
00:04:43
Speaker
That is split across the African Development Bank. So um I haven't gone too far in that, but I'd imagine it's a bit like the European Central Bank. So it's it does financing agreements for various countries across the continent. yeah There is about one third is coming from private sector investments, and that's built up of energy manufacturing transport companies. so like those who are going to benefit massively from yeah the highways itself ah the member states so you've got the the five ah countries that it crosses yeah um so directly from those governments and then it does say there's some public private partnerships going on but I haven't got too much detail there it's it's some it' still some low percentage of actually committed funds is that right
00:05:25
Speaker
Yeah, I think there's only, it's less than half at the moment is confirmed or secured funds there. So um yeah, still looking a bit sort iffy in terms of will they get the final funding? Will they even start the project before they secure the rest of the funding?
00:05:39
Speaker
They might ah make a call there. Contractor wise. Yeah, so let's say it does go ahead. you know who yeah who who Who are the likely ah contractors that we could see bidding for and maybe winning the work?
00:05:51
Speaker
Yeah, so the the name that some of us may recognize from ah Europe and others this is Bessix. um So that's a contractor from Belgium. ah They do quite a few infrastructure jobs across the continent.
00:06:03
Speaker
um So that's a ah likely contender. ah Then you've got a couple of the big Chinese companies. um So the China Road and Bridge Corporation and China Harbour Corporation. ah China Harbor Engineering Company.
00:06:15
Speaker
um There's a couple of local or on-continent contractors, so the Great West Africa and then Mota Engel, not contractors that familiar with, but have a track record of delivering infrastructure. It's amazing how many of the large European contractors have got a background and history of work in in Africa. Most of the the well-known names actually.
00:06:38
Speaker
So Vinci, Afarge, WeBuild have recently built a big dam there. um I think ah even Athiona have some historically some projects there. So I think we you could see like some large ah European contractors bidding, ah probably will see bidding for work.
00:06:56
Speaker
There will be some complexities. So it looks like each country has different, not just standards or engineering standards, but even procurement laws and things like that. So it'll be pretty difficult for any contractor that has a section that spans multiple countries and trying to manage standards and and laws on either side.
00:07:14
Speaker
Yeah, you wonder how they're actually going to split the contracts. It would make sense to have like five, across the five countries so you're working within the same set set of standards within a team gets super complicated if that starts to change within yeah a single section But yeah, it feels like it still feels pretty like hopeful at the moment given the funding situation, but it's a it's a very ambitious project. Like $15 billion USD project ah in ah in a more developed... like that That's probably a developing market. That's probably, like I don't know, like if you were to deliver that same amount of work in ah in a developed market, it's probably going to cost four or five times as much or something like that. Yeah. The timeline and budget is the...
00:07:54
Speaker
the The shaky thing at the moment is a lot of noise in the media. So want keep an eye on. Yeah, definitely project to keep an eye on and see if it does get approved and who ends up winning a bunch of work there. Let's move on to a really interesting set news stories that have been coming out recently about different construction contractors building tools internally thanks to AI.

AI in Construction: Build vs Buy Debate

00:08:20
Speaker
um So there are a couple of different stories recently with different contractors like Skanska and another one, Granite Construction, that built internal tools um off the ah using AI.
00:08:34
Speaker
So one was a a tool called Ashfelt Guru, um which ah helps basically internal knowledge sharing or um around Ashfelt inside of the business.
00:08:47
Speaker
And then another one, which was called a Safety Sidekick. It does spark a conversation or a threat a trend of like this idea that we've talked about before about building versus buying software or technology as a construction contractor.
00:09:02
Speaker
You know, I guess the question is, does ah AI ah remove barriers to entry for building software? um through things like this vibe coding trend ah where you can just talk to AI and have it build applications for you.
00:09:20
Speaker
Does that ah totally change the balance of like what a contractor can build versus what they should be thinking about buying? Yeah, I think to start with, um there's a lot of ah contracts and teams that I speak to that um they want the we've got all of this information and we want to be able to query it at lightning speed and get useful information out the front end and all of them immediately like if they look at a tool they go we don't want the data going with x con like software from the us or whatever so you can see why specifically with these two two examples it ticks the box of like they're holding it in-house um so like it it makes a lot of sense um
00:10:03
Speaker
to go for tools like this. If we look at the, obviously there's the advantages that we've spoken about um ah bunch of times, which is like they can highly customize the tool to whatever they want to do because it's in house.
00:10:15
Speaker
This, it removes the like the cost, the third party involvement, the, if you work with a software vendor, you can't just get everything that you want on day one. so you could see why it's a tempting thing to do in-house.
00:10:28
Speaker
With a tool that's like generating a response, and we're gonna be talking like at the outcome of this service later, like that seems um ah with my limited technical knowledge, something that is like a bit more manageable in terms of having a team internally, developers, design, maintaining, at all the cost that comes with sort of building, owning and operating software. And then there's gonna be a whole bunch of tools in the SaaS market that are kind of becoming redundant because if they were just like dashboard focused tools using data, like that's a, you're going to be in a, yourre you're up, your competition are the people that you're actually trying to sell to who should and will know best.
00:11:09
Speaker
But it's, there's like a scale of complexity that I've got in my mind and I can't see it going too far down the complexity end because the cost of building, maintaining complex software, unless contractors are planning to sell it to other contractors, there's a huge amount of money to burn on what we, a huge, like you're going to need a whole bunch of tools if that's your sort of strategy moving forward. to So yeah, just to like push back on that, I guess one of the things that we would see, uh, contractors think about if they were doing this build versus buy decision three years ago
00:11:39
Speaker
is um The cost of building a tool internally is not just the cost of building it, but it's like this lawn that needs to be tended to. It's the maintenance of that software addressing bugs, improving the functionality. And then suddenly you've got this like fixed cost.
00:11:53
Speaker
If AI today could, ah you know, we've seen this example of Claude, the AI tool have this you know app that they have called Claude Code. That app is 95% built by AI.
00:12:06
Speaker
ah So if if the if the capability today is at least some degree of um reducing cost of maintaining the software by ah empowering and um supercharging existing development resources and you know fast forward a year or two years,
00:12:22
Speaker
it can like solve some of the bugs itself and ship the updates uh it does like fundamentally change the the the cost of building so you get does fundamentally i guess what if i was to push a little harder you take it to the limit maybe you take away the cost element as even a consideration at all as to why you would go and buy so at the moment there's a you would consider build versus buy the build has this upfront capital cost and this ongoing maintenance cost imagine that like cost totally disappeared how does the build versus buy equation start to look then i still think i get you can't
00:13:06
Speaker
My ah quantity to surveying mind with the idea of removing cost out of the equation makes me stressed. But like in the example you said there of the 95% with Claude, how much does that 5% cost? If we look at contractor margins, it's still gonna be a serious amount of money to not just like be the person that vibe codes it, checks it,
00:13:28
Speaker
update set, there's still a serious, like anything more than like a light form. Maybe if i if I was to give it a specific size. So let's say we were yeah we were looking at some workflow that we needed to solve inside of a construction contractor.
00:13:44
Speaker
And the um the equation that we used to we would used to be faced with would have been like, we can build this thing. We need like 3 million or 5 million or something right now. And there'll be some ongoing costs or we could spend 300 grand a year or 400 grand a year and buy this tool that does it but we you know we're locked into this vendor let's imagine that the capital cost now is we need 250 grand to build it and the external tool is 350 grand or something still yeah you know maybe we're not there now but maybe we could be there in a year yeah how yeah what what other is is that a ah decision made we just build it
00:14:21
Speaker
Oh, it's a difficult one to answer because on one side of the fence, you've got this like, we can do whatever we want. we can We can define it. We've got no other sort of third parties driving us in a different direction because they're trying to serve a market. On the other side, like, are you siloing yourself to potential because like,
00:14:39
Speaker
unless you back yourself is we've got all the knowledge in the world we don't need to know what other people are doing we don't need something that others are contributing towards it's maybe a bit of a arrogant view on the world that like the benefit of a tool that serves everyone is the collective feedback from everyone to to push it in a in a direction um so suddenly you're you're reliant on internal knowledge and if we we know contractors A lot of the best minds in a contractor are not providing feedback to tools in a central space. They're delivering projects and they're super busy.
00:15:12
Speaker
So you really have to back your ability to, to like gather, track, interrogate, and like have a really solid process to make sure that you don't like leaving it on the shelf to just sort of die. Yeah, yeah, to build something. that Yeah, because like you're right. like If you zoom and you think about the the cost benefit analysis of any piece of technology inside of a construction contractor, you have a couple of pieces that go into the the calculation or the business case. You have the ah potential benefit of the of whatever process or tool we're implementing.
00:15:48
Speaker
And then you have the cost of it, which is the actual subscription cost and then all of the associated implementation, like our internal resource cost, IT overhead, et cetera.
00:16:00
Speaker
And if we just focus on the cost side, usually the actual subscription cost is not the total. It's definitely not the total, but it's it's a minor part of the the cost equation.
00:16:12
Speaker
um So yeah, you're right. lot like there's like it's The subscription or the the buy thing is not the total story of of the cost of it. So you don't really lose. It's not a one for one.
00:16:23
Speaker
yeah and then And then on the benefit side is the thing that you were like probably prodding at. And if I play it back to you own words, is like you're backing yourself to build something. that is ah high quality enough that folks will use it and you and through using it, you can capture the benefit that you're expecting to gather because most software or technology purchase or build decisions are based on some multiple of ah ROI. Like we get three times back for what we spend or four times back for what we spend.
00:16:55
Speaker
And so if you fail to nail the ability to capture the benefit because you don't make it to the same quality or you can't build it in the same time or whatever the the tradeoff is, if you fail to capture the benefit, that's actually way more impactful than 100 grand on the cost or 50 grand on the cost or 300, you know.
00:17:13
Speaker
yeah So ability to execute and deliver the benefit is probably the most important part of the whole equation, more so than the cost. Yeah. And also, let's not forget, when we got the complexity scale, even doing things like support and onboarding, like, are you going to have teams to do that? And then on top of that, the clients are going to expect the same standards that software vendors have to adhere to, like so certifications.
00:17:37
Speaker
yeah You've then got teams having to manage that. And like a software house gets the benefit of you've got one cert and it serves every contractor. That costs spirals when everyone's doing it individually. Yeah. And then like complex workflows, you probably get to some result where like we probably won't be able to build it and capture the benefit. And so you end up with the result probably of the contractors building the kind of small workflow tool things.
00:18:01
Speaker
But we have also seen a constant refrain and conversation from contractors saying, we hate lots of point solutions. We want one thing to solve every problem. yeah And the result of this path probably gets you to like, yeah, we've got 75 built inbuilt tools that do, how you know, here's the guru for asphalt.
00:18:20
Speaker
Then there's the guru for concrete. And then there's, you know, you you suddenly have like 700 apps that are all these internally built things that are like questionably maintained. Yeah, then in come the consultants to connect them all later because they realize that they're sort of standalone.
00:18:35
Speaker
Yeah, so it fairly it actually sounds like I'm being negative on the thing. I think it's really good that they can build it. I think- 100% it should be done for company-specific or simple workflows rather than buying terrible tools that don't deliver that much value. But because- We couldn't build it before.
00:18:54
Speaker
If we can build it now, great. Yeah, that that that that complexity scale can't be ignored. We know from like construction ourselves, people love to build like macro spreadsheets. So this is like that on the heat. So you can see there's gonna a massive cohort of people really excited to go and build everything. And you all have- Did you that on heat, by the way? just i'll just pause it. I've never heard. I would have said like that on steroids or that on- Yeah. That on heat. Same thing.
00:19:19
Speaker
yeah Yeah. But you can imagine- um Unless you're really ah structured and have some guardrails up, you're going to end up with like six versions of the same tool in the same contractor when people are getting their hands on.
00:19:32
Speaker
Yeah. Yeah. It becomes the proverbial Excel spreadsheet. Yeah. Yeah. Yeah. So um yeah. yeah no that's Yeah, that's true. So i yeah, I think. But what is really good is it definitely pushes and pushes it pushes construction software vendors because at the end of the day, there's this yeah is this trope or there's not trope. There's this like heuristic, which is if you want to be a vendor that gets product market fit.
00:19:59
Speaker
you need to be 10 times better than the status quo or the alternative. And if the bar just lifted a little bit for what the alternative or the status c quo is, it pushes pressure on the vendor to be better, solve a harder problem, yeah yeah build a better product. And I said think that's great. You have to be better and and that bar's racing at quite a rate. You have to be.

Impacts on Venture Capital and Market Dynamics

00:20:20
Speaker
has other interesting side effects for like the venture investing dynamics. yeah Yeah. What's the clock on this? Because you know we talked about we talked about in the previous in a previous episode, this idea of the the venture capital coming into construction technology works on this basis of ah this power law dynamic where some small number of the investments massively succeed.
00:20:46
Speaker
Some of them in the middle basically break even and then a whole bunch of them might fail. And so ah it's the responsibility of the ones that succeed to return so much money to the investor that it pays for all of that times two, three, four, five times ultimately whatever the fund ah needs to return.
00:21:06
Speaker
yeah um which then puts the pressure on that software company to, through its fees that it charges the construction company, ah extract that enough value from the construction company to pay that back. And now suddenly if the the bar keeps lifting and the and the build versus buy equation puts pressure on pricing,
00:21:29
Speaker
um it does change the dynamics a lot. Like if if if these companies that have, you know, had $50 million, $100 million, dollars you know, raises recently of $60, $70 million, dollars you have to you have to build a massive business that is making hundreds of millions in revenue to to return that investment and be one of those winners.
00:21:50
Speaker
And suddenly there's pricing pressure. ah coming from the ability to build internally. I think that is a really good thing for contractors. It means that contractors are not stuck taking whatever rate ah the the yeah the winning software companies choose.
00:22:07
Speaker
If you imagine a world though, that um let's just say that contractors build way more than we thought they were going to build. and like really like nestle in and like invest in teams that their job is to like build these tools in-house. yeah Do you think it would slow down the development of the whole market because building tools for one contractor, your budget is so restricted because you haven't got the the the income from others.
00:22:30
Speaker
so do you think it actually slows down everyone even though they all think they're moving faster? Yeah, you get this kind of like ah almost like this prisoner's dilemma thing where everyone can say this is better for me because I can use my internal, my intellectual property that I have as, you know, contractor A or B to build the tool that works best for Yeah.
00:22:53
Speaker
yeah But you probably at a macro level ah result in, you know, many, many people solving the same problem many different ways, but no way to scale that um to benefit. You know, if someone does it well,
00:23:08
Speaker
you don't get that ah efficiency benefit scaling across the whole um the whole marketplace. So you probably don't impact productivity as Yeah, you probably don't see like market-wide or industry-wide productivity improvement at the same rate by having by having like the knowledge locked into one contractor.
00:23:27
Speaker
Yeah. And also you're gonna probably invest, let's say you take that option, you're probably gonna invest in a team and they're gonna go through like projects When in reality, you need a team for each to keep leveling them up as you go. Otherwise you end up with these sort of, you're always on the next thing that needs to be built and letting the other ones sit. So you wonder how, yeah, you can imagine that the pool of resource required is just gonna grow and grow and grow.
00:23:53
Speaker
And then if suddenly you go, oh shit, we've all these people and actually, We're not doing this very well. Yeah. What's true is there's definitely going to be a space. There's going to be ah a definite economic space for the buy decision where it's like, there's just clear economic benefit to sell it to building some backbone infrastructure tool that everyone uses. It's like, ah yeah. Yeah.
00:24:15
Speaker
um But no, it's a very healthy competition. and I think it pushes the the bar raises the bar for sure. Yeah, totally. All right, let's move on before we run out time because there's a really interesting ah third topic where we have seen a lot of folks recently in the LinkedInosphere talking about outcomes as

Outcomes as a Service in Construction

00:24:37
Speaker
a service. And if you're really cool and a know you shortened it to OAS,
00:24:42
Speaker
um And so ah what outcomes service is if I was to like boil it down, ah you can think about how you would typically buy technology or software in construction ah right now, which is that you might pay per project or you might pay as per percentage of revenue or turnover of the business or you might pay per user to get access to some tool. But that tool or that solution or that workflow isn't guaranteeing some outcome, it's just access to some technology.
00:25:10
Speaker
And the idea of outcome as a service is that the pricing model changes so that it isn't percentage of turnover or project value per seat. It is per result or per outcome that is being generated through this technology. Um, And this has been, i would say, popularized probably because the pioneer area where AI is really taking hold, one of the areas that AI is really taking hold right now is in this customer service space.
00:25:41
Speaker
So as ah as ah an example of of this transition from ah legacy per seat or, or, uh, access pricing to outcomes of the service.
00:25:52
Speaker
You can think of the, uh, a piece of software that does customer support technology. So you use some application, you chat to the, you chat in a little chat bot, someone on the other end answers your question.
00:26:05
Speaker
Um, and. ah you you get on with your day using the tool. Historically, that software would have been priced per user for the company that is you know responding to those questions.
00:26:18
Speaker
Recently, this market has changed a lot where you can start to get AI to answer those questions. And the way the pricing model for most of those work is on a per resolution basis.
00:26:29
Speaker
And so you don't pay anything for the software access, but you do pay on a per every time that resolves a user or customer's question, you pay per resolution.
00:26:39
Speaker
And so this has some, this idea about comes a service has some really clear benefits in, in construction, um, or in, uh, to any, um,
00:26:50
Speaker
customer because you're paying for a result but i think there are also some trade-offs there as well so i don't know what are your thoughts on the idea and this trend of our outcome as a service if we think about the the sas space there's definitely a lot of tools that you'd immediately have buyer's remorse it's not really delivering the value that you thought you would it's an absolute ball lake you're trying to train everyone no one's using it and you remove that whole sunken cost aspect if you are paying yeah for the outcome.
00:27:18
Speaker
but On a resolution sounds like a really gray area. Like who's to define whether that question answered or like output. Yeah, fine. That's the thing that we're paying for.
00:27:29
Speaker
That must be a huge gray area. So i think we'd you'd have to kind of think about the specific use cases to really think like how difficult is it to not just measure value. Yeah, what is an outcome given the specific thing? Yeah, 100%. Yeah, so if it's like a copilot type thing and you ask a legal question and it gives a legal answer, maybe that's an easier one to measure, but...
00:27:52
Speaker
If you want something that's slightly ambiguous or or subjective, like what are you paying? Is it just per answer, whether it's right, wrong or somewhere in the middle? Yeah. So to to like, to pull on that thread of like, what is an outcome? Uh, even in you would think that the idea of like resolving someone's question would be on the scale of the things that construction has to do, like a relatively simple, uh,
00:28:16
Speaker
measure of outcome. ah But even in that scenario, you kind of get this experience where you ask the question, it answers the question, and then it asks you, did this resolve your question?
00:28:27
Speaker
Now, the person asking that question say no ah do one of two one of three things. They can say, yes, it does, whether it does or it doesn't, because they could be kind of like they feel guilty or like they might be letting the company down if they click no.
00:28:43
Speaker
and they or they might think, oh, if I say no, they're going to ask me some more questions and I don't want to answer any more questions, so I'll just say yes. yeah Then you've got the scenario of like you say no and then you could be doing that because you didn't get the answer. You could be doing that because you you don't want to you know you don't want any follow on question or whatever. Yeah, or pay. But yeah, the customer is not paying for it. Usually it's the person providing the service.
00:29:08
Speaker
But then you've got this third, which is what I would probably do is I just close the browser. And leave. I'm not answering to some stupid bot thing that wants to know whether I whether i got an answer or not.
00:29:20
Speaker
And so in that scenario, you know what what happens? And that would be a relatively simple example of like what could happen what could happen with a relatively simple example of like did I answer your question?
00:29:33
Speaker
You use the example like getting some legal advice or whatever. It's like, well, did you get advice? Was it good? Did you get sued three years later? Because it was total bullshit. So yeah, I think outcomes of the services. Yeah, exactly what you've pointed, which is like, what is an outcome and how do you measure the outcome?
00:29:51
Speaker
is you could easily think of many situations where that the definition of what that is is very, very, very hard to measure. yeah What's good about it is, as you said, it like there's all these risks that typically live on the customer right now where I buy some software,
00:30:08
Speaker
I've got implementation risk. I've got people using at risk. i'm I'm taking risk as to whether over time this delivers the value that I'm taking. And theoretically, what happens with outcomes the service is it pushes that risk back to the vendor of like, you have to deliver these outcomes.
00:30:22
Speaker
But if you can't divide- Yeah, there's no there's no hiding, right? Because there's a lot of promised value or outcome from a tool. Yeah. um And it may or may not get there. And that could be a user behavior thing. It could be a limitation of the tool and that that disappears. So- Like in theory, that's a really good spot to be that you're only going to pay if you get the thing that you actually need.
00:30:43
Speaker
but but But what's interesting here is what we've just talked about is this idea of like the risk moves from the customer back to the vendor and risk allocation or risk transfer is ah very thing a thing that we're very familiar with in construction.
00:30:58
Speaker
And if we think about what happens when risk moves around is usually we price accordingly. Right. So like if you want me to take on the risk, then the price of the thing is going to probably ah go up accordingly. And so, okay, then you take the scenario of like maybe this is wildly successful and helpful. It helps me on the business. It could be that now by paying outcomes of the service, I'm paying five times more than I would have paid if I just bought the software on a per seat basis because yeah of the risk that has been transferred back to the vendor.
00:31:30
Speaker
But also, how do you measure? like if If you think about that the the scale of what a question and and answer could be, for example, in construction, Like, is it like word count? Is it how long the tool- It's build a build a wall. Like you could have an outcome, which is build a wall or per brick laid or like some outcome that's like you could, you tie it to like your unit of measure and inside of a car, like what you would pay a subbie or how you would pay a QS or you might think about like quality inspections performed. So I think you could think of these- But it needs to be something that you could like forecast how many of those events are going to happen during a job. Otherwise you have no idea what you're going to spend. But you take any of You get like shit questions from junior staff and you're like, stop it.
00:32:15
Speaker
ah Imagine you had this AI robot that could walk around and do your quality inspections. are you incentivizing it to how many inspections it does or how many it passes or how many issues it like uncovers because you get in this weird spot where like suddenly it's better off being really like negative and nitpicky on its quality inspections because it gets steep more and more yeah yeah yeah exactly yeah i'll fail yeah So yeah, there's like the the outcomes. The outcome thing is you can see how it would get like really like funky and skewed.
00:32:46
Speaker
Yeah. I think there's this principle in construction where we, you know, when we're negotiating or tendering a project where it's like risk should sit with the party that is best placed to manage the risk.
00:32:58
Speaker
And so if you push everything back to the vendor and you say, i will pay you per outcome, but you don't know about the specifics of my site, you don't know about the specifics of how we operate around our project, suddenly that vendor is like taking on a bunch of this risk that they can't really control.
00:33:14
Speaker
Yeah. And I think that results probably in like high prices or big ah You end up in like a world of insurance where they do price the risk and then you're overpaying because they don't know who you are and how you drive.
00:33:25
Speaker
so So I think that like the benefit that I would see for outcomes of service, which is at a very surface level, like you pay for a result and you minimize the risk of like I deploy some technology, I don't get the benefit.
00:33:37
Speaker
At the limit, it's either not sick like I think you either end up paying more and And we already have a mechanism around how we control that risk as a construction contractor, which is we pilot it and we test it on two projects or on five projects. Like if you if you think about the topic of outcomes of service for two minutes, you probably get to the result of this is a really good thing as a construction contractor. But i think if you think about it for 10 minutes more, you probably get to a spot where it's like,
00:34:04
Speaker
I don't know if it helps in many situations. Yeah, because whoever the the vendor is providing the service, like it only works because you're doing it at scale and construction always goes back to his everything that's like unique about this project or this this situation. So yeah, I think it'd be quite difficult.
00:34:21
Speaker
Like again, there's going to be that complexity scale. It's probably going to be awesome at some point. something uh or like a series of of things that are like the low complexity like high yeah like let's say like let's say there was some commercial software that monitored all the activity of something that happens on a project and identifies potential early warnings and it would just ping up and say hey based on legitimate ones and you go yes this is a this is a potential early warning that so you pay for that like Yeah, you would.
00:34:50
Speaker
Yeah, yeah. So you can see how some scenarios it makes sense. But I would fucking hate you if you had a something raising early warnings. Yeah, well, that's all right. Because they would buy they would buy the opposite software. They respond, yeah. that Yeah, that like takes in the early warnings and drafts responses and you click OK and you pay ah message back the thing. It's probably the same vendor doing both sides as well. But yeah, you have to probably ask yourself a meta question of like, is that an outcome or is that like just another step in the chain? Because like really is the outcome risk mitigation or progress?
00:35:25
Speaker
So yeah, it's it's it's interesting of what the definition of an outcome is and how you would price an outcome. Because like, yeah, what is the benefit to a contractor of early warning issued per early warning? Yeah, you probably said the other example but like.
00:35:41
Speaker
Yeah, the other example that you said, like makes a lot of sense, like a quality inspection being signed off. That is a like if it's ah and ah in accordance with the works information or the process that you have to follow and it's and it's done.
00:35:54
Speaker
that feels like a thing that's like now complete. Whereas an early warning might not go anywhere. It could have been like the most pointless early warning that was ever raised and like, why are we raising it? but But also like you said, you made this comment as like a QS, you go, I'd pay five quid for that, right? Yeah, just on the off chance, they're picking up one that like you yeah i know but like let's let's do the math on that let's say you were let's say you were issuing and you're paying five quid what's the average like a hundred million dollar project how many early ratings per month you reckon you'd be issuing 30 maximum probably 30 upstream probably three or four times that downstream okay so let' let's say let's say we said a hundred a month that we're issuing yeah uh
00:36:33
Speaker
So, well suddenly per early warning you're gonna pay, so you're gonna 100 early warnings times five a month, which is 500 bucks a month. Yeah, five. Which is six grand a year. That's not enough money to run a business on.
00:36:46
Speaker
if you're earning it If your average contract value, so if you're the software vendor that's selling this now, you're getting six grand a year per project only if they're issuing a hundred early warnings a month. I think the price of the early warning needs to be like 25 quid or 50 quid. Like it would be a lot higher. Oh, a hundred percent. Would you pay? If it was five, you would.
00:37:07
Speaker
But if it was a hundred, you'd probably be like, oh, that's worth if it's right. But for all the mistakes, like, is it worth it like stick someone on it. Yeah. You would get to that. Would you would you pay? Yeah, exactly. Would you pay 50 quid a month or 50 quid per early warning?
00:37:24
Speaker
Per early warning that results in a compensation event, maybe. Oh, but then like the vendor has no idea if it. And yeah, exactly. The vendor has zero idea. Like if you could measure to it results in a compensation event, you'd pay, you'd pay probably like 30% of the value of the compensation event.
00:37:42
Speaker
Like if it identifi like you you pay some meaningful percentage, but you can't measure that through. So, yeah, it's it's I think it's a very I think like I said before, it's like I think if you think about the idea of outcome of service for like 30 seconds or a minute, you go, oh, brilliant, you know, pay for what you get.
00:38:00
Speaker
But I think when you think about it for a bit longer, it's you can see it's like very messy. This is it super interesting topic. ah I know we're super over time.

Closing and Engagement Encouragement

00:38:09
Speaker
i be um We should revisit that once we see like some decent examples in industry, which we should see popping up. But um yeah, we are way over time. So um thank you very much, everyone, for tuning into today's show.
00:38:23
Speaker
If you did enjoy the episode, please do think about liking this video or following us on your chosen podcast platform. We really appreciate your support and we'll catch you all next week. Bye-bye.