David Whitcombe: Exit Prep Is Now AI Readiness - Why PE Needs to Start Earlier

Speaker A: Hello and uh, welcome to the AI Pathfinder for Private Equity podcast. AI Pathfinder helps private equity firms make sense of AI and make it work. It's an expert network built on insight, experience and connection. We bring the right people and ideas together to turn AI from something you're curious about into something that delivers real results. If you'd like to attend one of my regular AI strategy briefings in London, Manchester and soon New York, please check out the Show Notes for details. In this series, I've been speaking with founders and specialists who are helping private equity firms and their portfolio companies apply AI in practical, commercially relevant ways. Today's guest is David Whitcomb, Founder and MD at UH Data uh Vision Services. David has a background spanning analytics, strategy consulting and data driven value creation, including time at IBM and ocnc, and now works with private equity backed businesses to help them use data more effectively to drive growth. What I found particularly interesting about David's perspective is that he is less focused on AI as a standalone tool and more on what it means for workflow, decision making and commercial performance. In other words, how AI starts to move from being a chatbot on the side of the desk to something more embedded in how businesses actually operate. So today we're going to talk about what that looks like in practice, where the revenue retention opportunities are, how AI changes the way insights get used across a business, and what private equity leaders need to understand about the operational and commercial shifts now underway. David, welcome to the podcast.

Speaker B: Hi Steve, good to see you.

Speaker A: Thanks for having me, really m looking forward to the conversation. Let's kick off with just giving you a bit of a brief introduction, uh, to yourself and to uh, data vision services and also why stop this business and why focus on private equity?

Speaker B: Yeah, good question. So you know, as you said, my career kind of three jobs so far. I started very much as a data scientist data engineer with IBM. I then spent four or five years with ocnc, spending a lot of time around the private equity ecosystem, learning around diligence, learning around what good kind of framework led strategic thinking delivers. I set up DVs partly from a kind of personal perspective of I'd always wanted to kind of run a business. I come from a family of entrepreneurs and it was probably something written in the stars for me from a fairly early age and partly from an opportunity perspective in terms of believing that mid market businesses, 30 to 200 million revenue, don't have access to that combination of people who can get hold of data and go beyond Excel. Don't get scared when it goes beyond a million rows, but also talk to the board and understand what value creation really looks like. So I started in 2021, five years in ChatGPT, 3.5, came out one year in. So we've been kind of riding the journey alongside our, uh, private equity clients ever since.

Speaker A: Fantastic. Thank you. Well, let's get into it and, you know, we've spoken a couple of times and you've given a good sort of overview. So there's some things I'm going to pick up on those conversations we've had. Let's start with where we currently are today. And let's be honest, a lot of businesses and people are still using eight AI, mainly as the chatbot or assistant. What does it take to move beyond that and start embedding AI into actual workflows or even treating it more like a digital employee?

Speaker B: Yeah, I think there are two parts to that question. You've got the business perspective and you've got the person perspective. Now, from a person perspective, this is a huge mindset shift. A chatbot and an assistant helps you do what you were previously doing better and faster in order to move into agentic or move into full kind of copilots. It is a transformation of what you are doing as an individual. You know, you are no longer somebody who writes slides or builds Excel models. You are somebody who solves problems. And the, uh, interface that you have into that is now your favorite AI tool rather than cell E17 of your financial model. So from a personal perspective, it is a huge kind of transformation in what you're trying to achieve. And then from an organizational perspective, it's therefore about starting to think about whether these co pilots are real members of your team. You may have people whose previous role was more of an individual contributor role, but now they're managing a team of agents. So everybody's job titles within your teams are changing and you need to be able to communicate what that means at your team level.

Speaker A: Is that a realistic ambition in the short term? Um, or is there a lot of work that needs to get done to

Speaker B: actually realize that the definition of short term changes on us really quickly at the moment? Doesn't it usually? I would say no. Usually I believe that you have a series of early adopters, you know, 5%, 10% of the population of any kind of working population, and then it takes a long time for that to kind of roll out across others. I think there are two reasons why I'm very kind of AI bullish right now. I think reason number one is people are already using these tools independently in their own life. Right. ChatGPT is the biggest kind of personal use app. It's overtaken Google in terms of the way people run their day to day existence outside of work. So you've already got an awareness of what the toolkit is likely to be and secondly the leverage of that 5, 10% of people who are early adopters. If you do believe that they are able to take risk, their agents into being a member of their team, the leverage of that group of people will be far higher than they're able to generate through just normal societal leverage and normal people to people bonds. So I'm fairly AI bullish at the moment. There is a lot of work to do to get there. It may happen very quickly again.

Speaker A: Yeah, well let's just focus on that a little bit. What tends to get in the way when companies try to make that leap?

Speaker B: Yeah, I think people, data, uh, governance. I think we've explored people to a reasonable extent. I think data uh, is the critical piece here. And we're no longer talking about well structured databases. Right. We're not talking about a data platform. We are talking about all of the unstructured information in all of its formats. We are talking about the availability of SOPs, we are talking about the information in the corpus that you can build for your agentic tools inside your organization. I think there is a lot of investment that is going to need to be made in each of those over the course of the next 6, 12, 18 months to really charge the kind of next stage of the AI revolution. And then the other is governance. How do you create the right safety boundaries that enable people to innovate in a safe space without creating um, cyber risks, um injection risks or others into the organization? At DVS we kind of think of governance on two scales. For our own internal reporting everything is fully agentic. HubSpot talks directly to Claude. We use an ERP called Productive that connects to a zapier workflow. Um, we're still working on whether Xero is going to play nicely or whether we're going to move to QuickBooks just because it's the only one that will kind of connect well to our AI tools. I wouldn't dream of doing that with our client data where we're working with their CRMs and their ERPs. So where can you create safe governance boundaries for innovation? And where do you need governance to be a leader, not a follower?

Speaker A: Is there a particular, sorry, going back to, I suppose to the more people aspect, particular culture that um, means that There are or leads to companies moving further forward than others, I think so

Speaker B: I think there are certain incentives that you can bring about across an organization that enable businesses to move faster. And I think it is all about culture because I think you do need everyone to come along the journey. You need innovation to come from all areas. You know, had a lot of conversations over the course of the last 12 months how to create that innovative culture. You know, where is the carrot, where is the stick and um, where is the opportunity? You know, certain businesses, the Google approach for 10 years has always been four hours a week to do something outside of your day job and innovate. You know, there are certain businesses that are doing that well. Other businesses, ah, are kind of more doing hackathons. Other businesses are using champions and examples. But yeah, how do you communicate and propagate innovation once it happens and how do you incentivise the innovation in the beginning?

Speaker A: Well that's um, a bit of a segue into private equity because really there's demand on a growth, isn't there demand on profitability. And with a, with a technology like this, there is now clearly a route to achieving those that. Well, I say clear. You know, there is examples now where there's a path to achieving those things. So there is this sort of momentum, uh, building with this group and I assume, you know, that's one reason why you work with PBAC businesses. And if we start to just delve into that a little bit more, are you seeing the strongest commercial opportunities right now for those businesses, um, especially around the revenue growth and the retention piece?

Speaker B: Yeah, absolutely. We've spent a long time with our businesses. You know, we spend our time around ARR, nrr, grr, right. What is revenue retention and how do we grow it? So we've worked with businesses on kind of three or four different areas here over the course of the last 12 months, you know, from agentic outreach. So being able to build your universe of contacts, um, and make sure that you have the right triggers in place. X company has just bought another company or Y company just announced a new kind of product launch. Making sure that you're kind of working on triggers in terms of your outreach. I think the biggest one is around kind of proactive churn identification. You know, I've spent the last 10 years of my career building churn models based on tiny amounts of signal, you know, has someone logged into a platform. Now those churn models can take every customer service interaction, every piece of usage data, uh, they can understand how different, uh, parts of the Platform and tool are being used. So around kind of proactive churn and then I think around how you implement pricing and pricing is an area that we might talk about more over the next 20 minutes. Pricing is going to change a huge amount in B2B in particular over the course of the next couple of years. But how do you understand where there are uh, price increase clauses hidden in contracts that you're not executing? Or how do you understand from the sentiment of your sales and CS calls what your packaging needs to look like going forward in order to maximize on your upsell and cross sell across your organization. So I think number one it's about increasing the effectiveness of your sales teams with agenda outreach and your sales teams. Number two it's about better, more proactive churn identification and intervention off the back of it. And number three it is about using the suite of structured and um, unstructured data to make better decisions around pricing and packaging.

Speaker A: Is there an area that's um, being underestimated in terms of what the potential is right now do you think?

Speaker B: I think most leaders I see either have a good understanding of the opportunities within their organization so the opportunities in sales or the opportunity is in churn or the opportunities in pricing or they have a good understanding of the sources of insight that AI enables them to get hold of now. So you know, be that call recordings, be that the email suites, be that the CRM, be that contracts in PDFs, I think often the bridge or the building out of that matrix is what's being underestimated. The opportunity to use data source X for use case Y I think is understated. So the opportunity to use call recordings to better set packaging because you can understand the themes with which customers describe the opportunity is something we've worked on with a couple of software businesses over the course of the last 12 months.

Speaker A: I, um, I have a background in product management and so we did a lot of work around pricing and packaging positioning and I don't obviously spend too much time other than on my own services. But how valuable the, the, the tools are now that could allow you to um, be much more sophisticated and ongoing in terms of your pricing and packaging.

Speaker B: As we think about pricing and packaging in private equity, there's 100 articles a day about the death of the seat. And if you sell to employees and employee counts are going to change, the price per seat model is going to need to change. I think that's just kind of one part of the story because private equity is not just about top line revenue. It's about revenue quality. We're working with an organization at the moment who has a decent usage based pricing model. But that means that their revenue changes every month, every quarter, uh, and that means that their valuation multiple is going to be hampered because of the way their pricing and packaging is structured. So I think in private equity in particular you've got a two dimensional question of what replaces the seat. Because you don't want to just move to a usage based pricing model. You need to understand how you can create recurring, reoccurring and stable revenue profiles linked to value and outcomes, but which drive exit value, which is fundamentally the piece that your audience are going to be most motivated by more than just revenue growth.

Speaker A: So let's just stick there for a moment on the outcome sort of base pricing is the technology helping us measure that more successfully, understand better how um, the end user is using a product

Speaker B: I've not seen or indeed even as a consultancy that spends a lot of time around pricing, I cannot say that DVs have yet done it exceptionally well. I think the technology is becoming ready to inform that. I think that there are probably two things in my view holding organizations or the organizations that we work with back from really nailing this, I think one is the data foundation that you need to bring together much deeper, uh, usage based insights on a live and tracked basis. And two is contracts last for a year or more and the world right now is very different to the world of March 2025. The agentic world is very different to the world of March 2025. So you need to put in place data foundations, build up that data corpus over the course of a period of months to years and then wait for your contract renewal cycles to occur. So I think the technology is becoming ready. I've personally not seen brilliant businesses be able to use a greater uh, data source and AI based modeling against it to really reset pricing and packaging strategies to a great extent.

Speaker A: Okay. Another area that um, you're helping to um, I suppose make. Well it's dashboards essentially. And um, that's one of the themes from our conversations we've spoken about. And it's trying to go beyond that. It's getting the insights into the hands of the people in terms of how do they act upon that, what does that look like in practice?

Speaker B: Dashboards are dead. Long live the dashboard. I think it's always going to be the kind of center of this. But if we think about PE backed service organizations, so lift manufacturers and maintainers, engineering businesses which create kind of high grade components, 10% of the organization sit at a laptop every day. 90% of the organization are uh, out and about, hands on. And actually as you think about the macro trends of AI, you know, probably blue collar work is going to have its time again and we're going to celebrate the people with kind of physical skills there. So for me it's about how do you get the information that's previously sat on my laptop that I need a maths degree to really understand and kind of zoom around into the hands of the workshop foreman, into the hands of the regional manager, into the hands of the traveling salesperson. So we've spent a lot of time recently looking at how you transform a dashboard into a podcast or how you transform a dashboard into a prompted AI based email send into people's inbox. And we're seeing that democratization of insights really powered by media change.

Speaker A: And is that, um, I assume that that uh, sort of, well, you know, the podcast or that information is, is about. It can be really personalized. So it can be. This is the sort of things that we should now be doing rather than just providing again the information without that action.

Speaker B: Exactly. I think my favorite was a piece of work with a manufacturing business. We built their reporting suite for the kind of finance team within which you can then start to see, you know, we're going to sell 50 of product A next month and 100 of product B next month. You can then disaggregate product A and B into components C and D, for example. And we had an example where we were able to show that the requirement for component D was max of the production line in the following month and the requirement for component C was lower than the capability of the production line. So the podcast that goes out to the workshop form and says, you need to be watching chain D all of the time this week because if chain D goes down, you've got a problem next month, get your proactive maintenance in place, make sure that you're checking in on all the health checks. If anything's going to Amber, get on it now because that's going to be critical to your outcomes next month. So really being able to take it from. Okay, cool. We can see the top level forecast into what does that mean for individuals across the organization and critically what action does that individual need to take differently because they're able to receive that information?

Speaker A: Um, and what has that ultimately led that manufacturer to achieve?

Speaker B: So this manufacturer used to have, I think it was kind of 20% of orders kind of delayed through the process due to, you know, bottlenecks in production line downtime, etc. Not significant delays and not kind of contractual implications. Um, but through the pilots that we were working with them on we were able to get that down to 8% so we were able to reduce machine downtime, um, we were able to increase throughput, um, we were therefore able to pull revenue of those orders forward and of course then create space in the production line for the next orders to come through. So this was really about revenue growth potential without needing to invest in additional capex to scale up the manufacturing capability

Speaker A: and just bring in the P firm um into sort of uh, focus a little bit here. Does that make the, is this making that reporting back to the P firm um easier and does that enable the P firm to understand, you know, where the opportunities or risks lie within each of uh, their portfolios?

Speaker B: Not seen it change yet. We were talking to a uh, fund even last week who broadly still get semi structured management accounts and they put them into one of the classic software for GP reporting and then democratizing that insight even across their organization is still challenging. I think the way we are seeing it for now is that reporting upwards is still to people with maths degrees and laptops and it is still at the kind of dashboarding level. I was pondering this kind of ahead of our conversation. Uh, be interesting to see how this changes diligence again going forward. If information in the hands of the GP increases and they have intimate details on likely upcoming churn, big wins coming up or lack thereof, buyers are going to insist on getting close to equal information across that. So no, I don't think we're there yet. We're close and as the current owners start to get more information, the level of work going through in diligence is going to increase as well.

Speaker A: Let's follow that through in terms of what we think or what you think private equity leaders must need to understand about AI and how it could change business models, pricing, even the quality of the recurring revenue.

Speaker B: Yeah, I think what's really interesting to me is still the spread within P.E. firms. Um, there are various which have big catalyzing programs out there, kind of top 10%, top 20% moving things very quickly forwards. And equally there are still a number of funds who, their biggest problem is getting their warehouse workers into the office more than 80% of the time. How to work across that full spread is going to be really interesting for the businesses at the top end. I think they are transforming their organizations very well and I think the question is about how is this going to transform exit Value it is about what is the new pricing model going to work to drive recurring revenue? How do we transform our uh, P and L for a world where we suddenly have this new variable cost called computer services? Businesses who are going to now not just be paying their consultants 60, 70, 80,000 pound loaded cost but those consultants are going to be using 30, 40, $50,000 of tokens a year. How do you transform the P and L as that works through? And then I think uh, at uh the kind of more hands on end of the market we are still in a world of pilot, we are still in a world of incentivize innovation, find opportunity and you don't need to roll that out across the entire organization. If you can increase the capability of one of your team by 50%, if you can um, increase the volume of site visits that you can get to by 30% there are lots of small things that you can be doing at that lower end and I think it's going to take a while before the implications kind of come into those more hands on organisations.

Speaker A: In terms of those organizations, where should they actually start? Is it workflow mapping, governance, the data foundations bit or something else?

Speaker B: For me it's workflow mapping. Find a small targeted opportunity, deliver on it and champion it. Get your teams incentivized. So for example within DVS recently we have taken a specific project type that's kind of a commercial diagnostic or a pre um VDD exercise. One of the team has built an agentix solution against it that can move two weeks of work down into two days. We're looking to really champion that and demonstrate it to our team as look, this is what you can do, this is what you can achieve. Here's the toolkit that we use to do it, here's the rest of the tools that are available to you to get that first example in and champion it. That's the kind of sharp end of the spear I think. Then beyond that get the data foundations in place, get the SOPs that get access to unstructured um data across the organization and make that accessible to the right people and then align your risk appetite. Is governance going to have to lead here or can it be a fast follower? Or more specifically can you find areas of the business where governance can be a fast follower because the risks are lower? And do you have areas of the business where governance needs to be a leader and you need to make sure that you've got safeguards in place very early on? So what's the commercial risk of the data underneath it? What's the people risk, um, in different areas. How do you find the internal reporting that we're using at DVS that enables the brakes to be taken off?

Speaker A: We are approaching the end of our uh, conversation. It's sped by. But um, one final question here. Do you think some of the work that used to feel like exit prep, getting data organized, standardizing processes, cleaning up documentation is now really AI readiness work?

Speaker B: I think absolutely. I think private equity has realized over the course of the last three or four years that, that you need data to get through exit. I think if you're going to need to spend that money, do it earlier. Get your CRM integrated with your erp, get your call recordings all in one place. Use that to set your pricing strategy, use that to target your icp. Use that to make sure you're not leaking revenue through the contractual process every month or every year. Drive value, pay that investment off and then uh, go into the exit process demonstrating what a brilliant forward looking business you are. So yes, I think if your listeners take one thing from this talk today, Steve, it would be about exit prep should be happening earlier. And it's really AI readiness work.

Speaker A: Have you seen that sort of work through yet or is it too early still?

Speaker B: So only by accident. So we worked with a child software provider through an exit prep uh, 18 months or so ago that didn't end up happening, but we had done all of the work and in particular I remember a awful week that two of my associates had to really unpick price and volume effect inside that organization. The reason for um, unpicking price and volume was in order to be able to show that there was still price headroom for a future buyer transaction, uh, didn't go through. That organization now has all of that information in one place and they have been able to go away and you know, really understand the different ladders of their pricing model, really understand the different use cases that their different clients and customers have. And they've been able to, they've been able to grow um, GRR by 6 percentage points over the course of the last 18 months primarily through kind of targeted pricing work. So I've only really seen it by accident. Um, I'm still kind of beating the drum on get your data prep done earlier. The opportunities are there.

Speaker A: Yeah, everybody should be getting onto their exit prep regardless of where they are in the cycle. Um, look David, that's been really fascinating, really thought provoking. Thank you so much. Um, if you'd like to know more about David and the work of Data Vision Services we'll include the relevant links in the show notes. But thank you so much, David, for your time.

Speaker B: Thanks very much, Steve. Good to see you.

Speaker A: That's it for the episode of aipath Fund of Private Equity. Thanks for listening and we'll see you next time.