Latest / Private Equity Data Guy / What Actually Breaks Private Equity Portfolio Companies

Transcript

What Actually Breaks Private Equity Portfolio Companies · Private Equity Data Guy

• 0:00
  If you've got a 30 person company and they don't have an up to date org chart, then that's usually indicative of vague and ambiguous decision making in the company.
• 0:11
  And fundamentally that's a people problem, that's not a tech problem.
• 0:15
  We do a lot of work bringing the unwritten, unacknowledged challenges of businesses out into the open and turning them into things that people are happy to talk about so that they can succeed.
• 0:26
  We want to give people the fishing rod and the skills and the ability to to carry these responsibilities forward and then keep an eye on that over time rather than just taking it off their hands and leaving them
• 0:37
  fundamentally less agile, less capable, less skilled in the process.
• 0:42
  Behind every value creation plan, there's a data problem nobody wants to talk about.
• 0:46
  Fragmented systems, metrics nobody trusts, and decisions made on gut feel dressed up as analysis.
• 0:52
  Welcome to the PE Data Guy.
• 0:53
  Each week host Graham Crawford talks to the operating partners, advisors and practitioners who are doing the work inside portfolio companies.
• 1:00
  If you care about what actually drives returns in the market, then you're in the right place.
• 1:05
  The PE Data Guy starts now.
• 1:08
  Rich Kershaw has spent a career on both sides of the technology table.
• 1:11
  First as a CTO and product leader and for the last several years of running technology and product due diligence for VC and PE investors through Lechi Kershaw.
• 1:20
  On hundreds of investment rounds for houses like Octopus Investments and Maven.
• 1:24
  He co founded the diligency consortium of DD specialists and is now building Billy a platform that monitors companies against best practice between transactions.
• 1:34
  What makes him the rare person to qualify to answer today's central question?
• 1:39
  What should software judge and what should never leave human hands?
• 1:43
  Very front of mind for everyone today.
• 1:44
  Welcome to the show Rich.
• 1:45
  It's great to have you on.
• 1:46
  Great to be here.
• 1:47
  Thanks for having me.
• 1:48
  And another, I always give credit to bawdy for this, but another bawdy intro I believe.
• 1:53
  Indeed yes.
• 1:54
  I think we've both been wing walking on the frontier jet of the AI boom.
• 2:02
  That was almost a mangled metaphor but I stand by it.
• 2:04
  I love it.
• 2:05
  It's interesting seeing how it's developing because I think that's where a lot of people's professional contact is with agentic visitor tools right now.
• 2:13
  And yeah, enjoying walking on that wing with you.
• 2:16
  People always take different winding roads into private equity.
• 2:18
  And you know, we just spoke off camera about a couple of your steps.
• 2:22
  Like what are some of the steps on the road that led you to becoming involved in the wonderful world of pe?
• 2:26
  I mean that's it?
• 2:27
  It's a good question.
• 2:28
  It was a circuitous route and I mean, I've always rather held that, you know, very few people end up taking the career path that they imagined when they spoke to careers guidance counselors or equivalents in their late teenage
• 2:39
  years.
• 2:41
  I took a very roundabout route through media production, then into, I guess a loosely adjacent career move into digital marketing agenc, and then got roped into an escalating sequence of CTO roles as a founding director over a number
• 2:58
  of startups over about 15 years.
• 3:00
  And there was a point in my career where I had been the founding CTO or ctpo, sitting on boards, being at director level, exited a couple of times where I thought to myself, well, I've had lots of contact
• 3:14
  on the startup side of things and I've worked with plenty of companies alongside blue chips and big enterprises and I think I've got an understanding of how they work, but I've never been inside.
• 3:23
  There is a cultural piece missing from my lexicon.
• 3:26
  And so I actively sought out a role with a larger organization, landed at Octopus Investments, which were at the time they were in the relatively early days of getting household name status, at least in the uk.
• 3:40
  They are now much more recognized.
• 3:41
  They were still expanding their tentacles.
• 3:43
  Yeah, they're the parent company of Octopus Energy, which everybody in the UK and I think increasingly the worldwide market knows pretty well.
• 3:50
  But at the time they were looking to be a little bit transformative in how they treated internal it.
• 3:56
  And so my experience in innovation and practical scaling of platforms and startup culture was something that they were actively looking for at the time.
• 4:05
  Unfortunately, that collided head on with the pandemic and exciting plans to do hugely innovative things took a little bit of a backseat at the point at which everybody was working out what this new remote work first plan looked
• 4:20
  like for everybody.
• 4:21
  And I took the leap and decided to go solo because I had been increasingly borrowed by the Octopus Ventures team.
• 4:26
  We happened to be on the same floor of the building and this is really all just happenstance and yeah, good.
• 4:31
  Opportunities, right place at the right time, right?
• 4:33
  Absolutely, yeah.
• 4:34
  And I started to do a little bit of work with the Ventures team.
• 4:37
  Discovered that I was actually, you know, pretty good at synthesizing the combination of tech assessments and business model and all of the stuff that I'd been doing at C suite level or sort of in parallel with things like
• 4:48
  architectural design and build.
• 4:49
  They liked having me in the room during investment committee debriefs, deciding on things like go no gos on investment projects.
• 4:56
  And yeah, I Decided to turn that into a business and with, you know, a few business development blind alleys, as there always are when one sets up a particular path.
• 5:06
  On consulting, we found a niche, brought a partner on a couple of years after I initially set up, of course now as you mentioned, we've got the diligency consortium, including us and peer professionals in other verticals.
• 5:21
  And we have found that our particular USP of combining together technical prowess and operational prowess with the ability to compare notes and look at the big picture is, I mean, above and beyond anything else.
• 5:34
  It's fun, you know, it is, it's genuinely interesting and very rewarding to go into companies at a point where they are, happily for us, very willing to be cooperative because they're going through an investment round.
• 5:45
  Of course, yeah, they want to be excited about their product and it's exciting and it is fun talking to people about their passion project.
• 5:53
  And we've at this point seen 250 plus companies go through these stages as well as doing strategic advisory for portfolio owners like venture capital companies and PE wrappers and that sort of thing, who've got companies that they've invested
• 6:08
  in and want to make them better, faster, leaner, more accelerated.
• 6:13
  Being able to talk to those people who are subject matter experts.
• 6:17
  Their thing about how we can take the lessons that we've seen all over the place over like 25 years worth of practicing professionally in one vertical or another and saying, look, these are the common strands, these are the
• 6:31
  things that we've spotted, these are the roadblocks we're coming up against.
• 6:34
  We're not going to tell you how to build a fish focused medtech company.
• 6:40
  That's your bag.
• 6:41
  But what we can do is we can talk to you about things like strategy and governance and how to build a sustainable tech platform and how to instill good culture in your team.
• 6:49
  And as lifelong tinkerer who likes to take things apart and see how they work, and that's led me down lots of different paths.
• 6:57
  The ability to take companies apart and see how they work is just an immensely engaging thing to be doing.
• 7:03
  Yeah, I love that.
• 7:04
  So much to pick up from that.
• 7:06
  And you know, even back to the start we talk about the career counselor.
• 7:09
  I don't think I've met any high school student who says I want to be in private equity.
• 7:16
  And you know, as a result, I think there are many different and interesting paths that people take to get there, which is why I always ask about it.
• 7:22
  And I love your example of a fish focused med spa as well.
• 7:25
  Does this, like this where people go in and get their feet nibbled by the fish?
• 7:28
  They do, actually, yes.
• 7:30
  I mean, that was, that was a slightly ridiculous example, but not that far off.
• 7:34
  Some of the businesses that we've looked at over the years.
• 7:36
  Yeah.
• 7:36
  I love it.
• 7:37
  I always wondered about this business.
• 7:38
  It's like, how long do the fish last?
• 7:39
  Do the fish just keep eating people's feet skin for their whole lifespan?
• 7:43
  Oh, it's, it's an interesting model.
• 7:45
  Is that their whole diet?
• 7:46
  What is the recommended daily allowance of, of, of to skin.
• 7:50
  Yeah, exactly.
• 7:52
  And if you're a smaller fish and you can't get to all the toes, like, what happens if there's a quiet day at the med spa?
• 7:56
  Do the fish go hungry?
• 7:59
  Maybe they've got spare fish.
• 8:01
  Spare feet, rather spare feet that they use.
• 8:05
  No.
• 8:05
  That's a chilling prospect.
• 8:07
  Well, I'll have to find a Fish Spot owner and we'll have a future episode.
• 8:11
  We'll do a deep dive into the, you know, the operational mechanics of the.
• 8:15
  I'll mind the network and see who we've got.
• 8:17
  Fish spot.
• 8:17
  Yeah, I'll look forward to it.
• 8:18
  Yeah, it's going to be the best episode ever.
• 8:20
  I'm looking forward to it.
• 8:21
  Yeah.
• 8:21
  Similarly enjoyed the.
• 8:22
  I think you talk about tinkering and seeing how the.
• 8:24
  Taking things apart and putting them back together again.
• 8:26
  I think I phrased the same thing slightly differently.
• 8:29
  I love understanding the inputs of a system and then almost x raying the black box of the business to understand how the input changes can affect the outputs at the other end.
• 8:42
  But it's the same joy that I think exists a lot in the type of work that we both do.
• 8:45
  Absolutely.
• 8:46
  Yeah.
• 8:46
  And I think I'm a, you know, for better or worse, you know, I was, I was raised by some, you know, some, some very, I think, principled, careful thinkers who, you know, I think instilled in me a sort
• 9:02
  of nascent sense of outrage when things aren't being done in a common sense way.
• 9:07
  And obviously, you know, through my growing up and maturing, I imagine I was probably not the easiest guy to work with when I was 21.
• 9:13
  You know, I think, I think the, the growing up process and getting comfortable in your professional role knocks a bit.
• 9:20
  Of that out of you.
• 9:20
  You learn diplomacy pretty fast, especially in.
• 9:22
  The world of investment.
• 9:23
  Very critical skills and bolts.
• 9:25
  Yeah.
• 9:25
  But I think there is still that underlying inner sense of outrage when I see companies which are, they are skilled and they are passionate and they are very good at a particular thing.
• 9:38
  But you can See them heading for some sort of pitfall or collapse or unanticipated choke point or bottleneck in their growth.
• 9:47
  And there is a big part of me and you know, and the partners in the group as well, we all share this as a common strand where, you know, if we see someone coming up against a challenge that
• 9:55
  they haven't spotted, you know, it's an, it's an unknown unknown or they just don't recognize it approaching, we want to be able to step in and say, hey, you know, there's a thing over there.
• 10:06
  You haven't spotted it, but we have, we've seen it before.
• 10:08
  And the good news is that there are solutions to it.
• 10:11
  You can avoid it, you can scale around it, you can put in place a little bit of strategy.
• 10:14
  You know, we do a lot of work bringing the unwritten, unacknowledged challenges of businesses out into the open and turning them into things that people are happy to talk about so that they can succeed.
• 10:26
  That often quoted stat that, you know, something like 80% of companies, startups fail in the first three years, that's just an.
• 10:35
  It's an immensely frustrating amount of human endeavor that goes into trying to bring really good ideas into the world and innovate and make people's lives better.
• 10:45
  Hopefully not every company is trying to do that, but certainly the majority, most.
• 10:49
  Are, or at least deliver value in some way.
• 10:51
  Even if.
• 10:52
  Absolutely, yeah.
• 10:52
  And if someone's put tens of thousands of hours into something more than that across a team, if as part of our job we can help them get over the easily avoided road bumps to the point where it's the
• 11:03
  product in the hands of people, or they're doing something interesting with medtech or sort of high street retail experiences and making people's lives a little bit easier or more comfortable or smoother, or research going back into the right
• 11:15
  places, then I think we're doing a good job.
• 11:18
  Yeah, I know, exactly, exactly.
• 11:20
  And as you mentioned before, you've kicked the tires on hundreds of technology platforms for investors.
• 11:25
  And there's a bit of a lean question maybe because I think I know the answer from my own experience to some extent.
• 11:29
  But how often is the technology or the technology choice actually the problem or the pitfall that's waiting?
• 11:35
  So this is a slightly dangerous answer to give given that a good deal of our customer base knows us as tech due diligence providers.
• 11:44
  Rarely, I would say it hasn't been particularly common for us to go into a situation and find that there's something fundamentally wrong with a company's.
• 11:54
  Technology fundamentally wrong is also, you know, it's difficult to turn that into a completely objective statement.
• 12:02
  The number of ways that a company can build a product, can implement some technology, can, you know, run a platform, are wide and varied.
• 12:11
  And lots of people in the industry with very vested interests in you choosing their platform or even just agreeing with them on principle are multitudinous.
• 12:22
  Generally speaking, what we look at when we're evaluating the technology in a startup that's about to get some money or is about to be acquired is what's the parity match between what they're trying to do and how they're
• 12:35
  doing it.
• 12:37
  I think I can probably count on the fingers of one hand the number of times we've come across companies where the report that we've given back to their potential stakeholders or their board, if they're just trying to turn
• 12:48
  things around, has been as explicit as this is a disaster and everything needs to be rewritten.
• 12:53
  That doesn't happen very often.
• 12:54
  What we do often find is that you have companies who made their choices work for them, and perhaps the things that they've chosen to do in tech, and they're not the choices that you would make.
• 13:06
  Now, maybe they started building some software 10 years ago and they have iteratively improved it over a period of time, but it's in need of a refresh or a replatform or something, and that's fine.
• 13:16
  That becomes an input.
• 13:18
  And one of the things that we sort of try to treat as our mantra is there are really no positive or negative pieces of information in a due diligence exercise.
• 13:30
  They're just signals.
• 13:31
  And if we come across someone who is a massive egotist, then we won't necessarily say, hey guys, that you've got a narcissist on your hands over here, because that's a value judgment and it doesn't necessarily have a direct
• 13:45
  bearing on whether or not the company will fail or succeed.
• 13:47
  What we might very well say is some of the personalities involved in this probably not going to be right for the next phase of growth.
• 13:54
  But we can work with you to reframe their role and possibly move them into an evangelist role.
• 13:59
  Because it could very well be that the personality that got the company to the point where they are right now was absolutely appropriate when they were going through that stage of being, you know, making their small fish in
• 14:10
  a big pond appear much larger and getting that credibility earlier.
• 14:14
  And now they need to be, you know, maybe they need to reframe that, be a little bit less executive and Move into a place where they're the person who continues doing that evangelism.
• 14:21
  And the same goes for tech.
• 14:22
  You know, the things which were suitable for you on day one, which got you to market quickly and allowed you to scale up and be really agile as you figured out what your ICP looked like and you got
• 14:32
  your product market fit and not necessarily the things which are going to take you through into a big enterprise push now that you realize that the 100,000 worldwide customers you thought you had are in fact actually 50 very
• 14:44
  large specialist customers who need a more large scale smoke enterprise fit.
• 14:50
  And that's fine.
• 14:51
  So much of the work that we do is about taking what I think a lot of people would pull a face out and consider to be a technology smell and frame it in terms that a private equity or
• 15:03
  a venture capital generalist will understand in terms of risk.
• 15:07
  Because if we see something that doesn't look right, then it can be considered to be a risk.
• 15:12
  But if the risk is handled properly, then that can actually turn into something which becomes a little bit of a defensive moat.
• 15:18
  If it's treated in the right way and it's treated as a learning opportunity.
• 15:21
  And that's an input to figuring out what you do next.
• 15:24
  Anyone else who comes along is just as likely to end up in the same trenches as you do and having to figure that out for themselves.
• 15:30
  And by saying, we've learned a thing, we're going to replatform and do it the way that we need to do it this time, that's a head start that you have over everybody else.
• 15:40
  We don't often see things that are obviously just bad, but we see plenty of things where a course correction is the appropriate thing.
• 15:47
  Yeah, I'm aligned with you.
• 15:49
  And that's where I was kind of leading with it.
• 15:50
  Like very, very rarely is the technology or the data platform that people have chosen is the issue.
• 15:57
  And if it is, then it's usually one of two types for me.
• 16:01
  Right.
• 16:01
  Either it's as you're describing, like what you've got now, you worked to get you where you are, but if you don't change it, if you double your business, you're going to have to double your team to do this
• 16:12
  because you'll hit the limitations of this which weren't on the horizon for you before, they're now on the horizon.
• 16:18
  And you need to think, you know, a few months or years ahead for the limitations of this arriving or the other one.
• 16:23
  And someone on a call yesterday used this phrase and it's kind of Stuck with me is using a bazooka to kill an ant.
• 16:31
  Where people like have actually overinvested in technology and platforms because they've, you know, they've heard the marketing lines and you know, I remain to be convinced that AI has made people better at building software, it's made people quicker
• 16:41
  at building software and it's made people better at marketing software.
• 16:44
  But I'm not sure it's made the software any better.
• 16:46
  In any case you get people with these sort of complex web of expensive platforms to do what is effectively a relatively simple task.
• 16:54
  So those, you know, those are the, the only two versions of that, that, that, that I come across.
• 16:59
  But yeah, more often it's misalignment people issues, lack of understanding like a shared understanding about processes or definitions or whatever it might be that that is, that is, is, is typically what we, typically what we come across.
• 17:15
  So the types of findings that you mentioned, some of it's relating to the people.
• 17:20
  The people and the org chart.
• 17:21
  Right.
• 17:21
  So what can an org chart tell you that a data room or a technology architecture might not be able to?
• 17:28
  Yeah, that's a good question.
• 17:29
  So I think I take it a level up from that as well.
• 17:34
  And I mean we deal with, I'd say we probably deal with three different types of source materials around org charts.
• 17:41
  We deal with high level org charts with very loose assignments and definitions of things like responsibility and reporting lines.
• 17:49
  We deal with very detailed ones which have got a lot of information on there and we deal with entirely absent org charts and all of those things are equally useful in their own way.
• 18:01
  Probably the thing which we always look to take away from the org chart question is given all the inputs that we've had and that can be obviously the documentation itself, level of detail and verbal briefings that we've had
• 18:16
  because we will always say to teams when we talk to an slt, talk us through your what's the shape of your team?
• 18:22
  It tells you an awful lot about how a company thinks about and considers governance.
• 18:26
  I suspect that some viewers, listeners to this might think, well that sounds quite a long way away from Tech DD.
• 18:34
  It really isn't so TechDD and to a certain extent all of the things that surround it are very much around basically value delivery.
• 18:44
  How do you create the value in your company?
• 18:47
  Techdd is really important because you're trying to figure out whether or not there's a good parity match between what the company is building and whether or not it actually meets the needs of the market and their customer base.
• 19:02
  But it's not a very useful signal outside of understanding what the operational process has been, to actually establish what that should look like.
• 19:09
  You know, if you haven't got good strategic product governance in place, you don't really understand how you're making decisions.
• 19:14
  There isn't like consistent structure for tracking decision making and recording decisions and assessing the value and the importance of the things you're planning on building.
• 19:25
  Then so much falls off the back.
• 19:28
  You know, you could be wasting the time of your engineering team and just as importantly, destroying all of the enthusiasm and confidence in their employer in highly paid knowledge industry workers who are incredibly valuable.
• 19:43
  You could be building the wrong thing and wasting money.
• 19:46
  Fairly straightforward.
• 19:47
  You could be addressing the wrong bit of the market and trying to build something for a target that doesn't really exist.
• 19:52
  But to pull it back all the way up to the org chart, that's sort of where it starts.
• 19:57
  Because if you, if a company doesn't have a really good idea of where authority sits and who key decision makers are and whether or not there are clean lines of ownership over decisions, strategy, delegation, then it almost always
• 20:15
  means that there's probably going to be either some overcomplication around decision making somewhere buried inside that company, or there's going to be a gap and things are falling through that gap.
• 20:26
  Decision making maybe?
• 20:27
  Absolutely.
• 20:28
  In common situations that we see off the back of that are things like long running and really important initiatives to build a new product feature so that you can access a particular bit of the market that you can't
• 20:38
  get to right now where the project is owned by a group of people and all of those people, if you ask them, think that someone else is currently responsible for doing that.
• 20:49
  And it drifts without having really clear org chart guides to say these people report to these people, these people report to these people, you know, and a tool that we use pretty frequently, if we're imposing this stuff on
• 21:02
  a company that doesn't have it and we've been asked to help out, it's just something really straightforward like a raci assignment, you know, and it is startling how many people are not aware that just off the shelf solutions,
• 21:14
  just shorthand for this stuff exists.
• 21:16
  It sounds very buzzwordy, but saying this thing that's important in your company, who's responsible for that?
• 21:22
  Just tell me who's responsible.
• 21:24
  That's not rocket science.
• 21:25
  And I think by wrapping it in the racy buzzword, people go, oh, management consultants.
• 21:29
  But actually it's just a matter of saying, well, someone ultimately has got to have the job of periodically Checking back with the team and saying, oh cool, you're on it, or you know, someone in the team working on
• 21:40
  it has got to know who to approach to say, the assumptions that we made about this thing are actually not accurate.
• 21:46
  Should we continue with it?
• 21:47
  Do we need to do more research?
• 21:48
  Should we talk to a client?
• 21:50
  And if you don't have that stuff in place, then it's pretty clear.
• 21:55
  Looking at an org chart and when you say what does your org chart look like?
• 22:00
  They either go, oh, we have a very clear idea of what that looks like, we just need to set it out and write it down.
• 22:06
  That doesn't really matter because we've got eight people and that's okay.
• 22:09
  That'll be a recommendation to investors to sort that out in the next, say three months, 90 day plan.
• 22:14
  But if you've got a 30 person company and they don't have an up to date org chart, then that's usually indicative of vague and ambiguous decision making in the company.
• 22:26
  And fundamentally that's a people problem, that's not a tech problem.
• 22:30
  That's going to result in at least some of the people who are critical to you making money turning up every Monday morning and being really unsure as to what their job is.
• 22:41
  What am I supposed to be doing today is the thing.
• 22:43
  I've just been sent by this person who is acting as if they are my line manager but doesn't appear that way in the hierarchy.
• 22:52
  Are they my boss?
• 22:53
  Should I listen to them?
• 22:54
  Am I doing the right thing?
• 22:55
  I'm not really sure what happens if I ignore it?
• 22:57
  Yeah, yeah, yeah.
• 22:58
  I've, you know, in, in a month I will have been here 12 for 12 months.
• 23:01
  Someone's probably going to want a performance review.
• 23:03
  Am I, am I going to have my job?
• 23:05
  Am I going to get a promotion?
• 23:06
  Am I going to get a job title change?
• 23:08
  What can I put on my cv?
• 23:09
  Should I be recruiting?
• 23:10
  You know, look, talking to recruiters and that sort of uncertainty and we've seen this time and time again, it bleeds confidence and enthusiasm for getting good work done from a team's culture and it's a really good indication that
• 23:26
  maybe that team has outgrown the model that they had before.
• 23:29
  And possibly it might point to the fact that whoever is sitting at the top is maybe not necessarily the right person to be at the top.
• 23:36
  Maybe you needed an evangelist to get to where you are right now, but actually what you really need is a super experienced operator who marshals the troops and keeps everything organised in the right way and makes sure that
• 23:47
  things are delegated down and projects are on track.
• 23:49
  And that's a very different shape of executive.
• 23:52
  Yeah, absolutely.
• 23:53
  And you know, I also hear the R rules about Raci as well.
• 23:56
  But it's amazing how many times that a column in particular, you get multiple people and then when you try and enforce that rule of like, no, no, no, it has to be one person.
• 24:05
  In fact, when I sit up with clients, I always say one of the things I request is like, you need to name a single point of contact.
• 24:10
  Who's responsible for this work on your side.
• 24:12
  Who is accountable for this work on your side.
• 24:14
  Because if it's more than one person, then it's nobody.
• 24:16
  I think we know this as well.
• 24:18
  Even if I could dive down into the detail and say, we're implementing data governance here, someone needs to be accountable for the data quality of this table.
• 24:26
  Oh, it's a shared accountability.
• 24:28
  I know immediately that means no one's doing it, right?
• 24:31
  Yeah, yeah.
• 24:33
  Everyone's responsibility equals no one's responsibility.
• 24:36
  Absolutely.
• 24:37
  A really good example actually is that there was a company who a good portion of the viewers of this will 100% be familiar with them.
• 24:47
  They were a mid sized, pretty well established company.
• 24:51
  I won't mention the industry, that might be a little bit of a giveaway, but they had 60 or so people, they were professionally very successful, they had a good reputation.
• 24:58
  We were brought in by their investors because they had not succeeded in delivering a single new product feature in three to six months.
• 25:07
  And this was in a team where they had upwards of sort of 50 people just working on product team.
• 25:13
  Yeah, absolutely.
• 25:15
  And we talked to them about the org charts and they gave us the org charts.
• 25:19
  And then within about 15 minutes of our conversations with the people in the mid level tier, it became abundantly clear that the org chart did not resemble reality at all.
• 25:28
  And the soft power relationships between people and who really had influence was very different from what was actually on the chart.
• 25:37
  And there were.
• 25:39
  The culture in that company was not something that would raise alarm bells.
• 25:44
  Everybody got on very well indeed.
• 25:46
  They were all super friendly.
• 25:47
  They had a culture of being very collaborative and highly cooperative.
• 25:51
  However, no one had any boundaries.
• 25:53
  And so a person working on one project, but who was one of many people on the accountable in the accountable column, in the theoretical race that they definitely did not have, would be included on every single stand up
• 26:10
  for every single project that they were being consulted around or they were ultimately partly accountable for.
• 26:18
  And so the majority of people in that company spent four or five hours every single morning in Stand ups that overran and then they would take lunch and then they come back and have a little check in with
• 26:28
  their team and they go home.
• 26:29
  So, you know, it can work both ways.
• 26:32
  So everyone was very well informed about what was happening.
• 26:35
  Absolutely.
• 26:35
  But because we, because we started at the, we started at the org chart level when we, we sort of got a gut instinct that something wasn't quite right and we sort of walked that path down.
• 26:45
  That was the opener that we started with when we talked to half a dozen people in the discovery process.
• 26:50
  And it took probably in aggregate, it maybe took 40 minutes for us to diagnose what the problem was.
• 26:57
  So it's just an immensely helpful tool and far more useful than it would have been to start from.
• 27:02
  Tell us about your code base or talk to us about how you do data storage.
• 27:06
  Yeah, yeah, indeed.
• 27:07
  It's.
• 27:08
  Again, I'm a big fan of in person and you know, all of my teams spend time in person with my clients and that's, that's the why.
• 27:13
  You get to the, you get to the why so, so much faster and you can feel it around the office for sure.
• 27:19
  I want to talk about AI and software again.
• 27:21
  I think we talked about this before.
• 27:23
  There's a lot of platforms out there claiming that they can almost automate due diligence entirely for private equity in a cookie cut away like all the way down to the bottom.
• 27:31
  Private equity that I know of is skeptical of it.
• 27:34
  I'm definitely skeptical of it.
• 27:35
  I think you are as well.
• 27:36
  Why do you think private equity won't buy diligence in a box?
• 27:39
  So I think we've seen a spectrum.
• 27:44
  One of the ways that we have noticed that the market's been changing in due diligence, I think in general over the course of the past couple of years is, I think we're midpoint on the curve right now.
• 27:55
  There is probably going to be a little bit of a reality check dip, I would imagine over the course of the next 12 months or so.
• 28:03
  Correction.
• 28:04
  But yeah, absolutely.
• 28:06
  Our experience has been that a couple of years ago we were being asked regularly by our clients, by our investment clients, how they should be using AI.
• 28:14
  And to be honest, the things that AI are, sorry, the things that LLMs are good at hasn't really changed very much over the last maybe 18 to 24 months.
• 28:25
  You know, the things which they were powerful at.
• 28:29
  So, you know, semantic comprehension and digesting large volumes of information and summarizing, they're still good at that.
• 28:38
  Performance has gone up in terms of speed, comparable cost per token has come down.
• 28:42
  So it's more coding capabilities definitely improved as well.
• 28:45
  Absolutely.
• 28:45
  Coding capabilities have evolved just so much.
• 28:48
  And the ability to do complex stuff around coding and specific problem solving, but in terms of loose comprehension and giving opinions, nothing really has changed very much.
• 28:58
  And we've been talking for the last few years really about how really careful guardrails around the prompts that you're giving to your LLMs, and putting careful boundaries and specification around what it is that you're looking to get out
• 29:10
  of an analysis is really key to having powerful and very valuable usage of LLMs.
• 29:16
  Things like Microsoft Copilot, where you sort of say, write me a document to do xyz, I mean, that's kind of fairly throwaway.
• 29:22
  It's also fairly expensive.
• 29:24
  And I think we're demonstrably seeing that an awful lot of companies who signed up for Microsoft Copilot in Office365 and GitHub and so forth, finding actually that sort of very piecemeal send a chat message and expect something quick
• 29:38
  back isn't really paying tangible measurements.
• 29:41
  They're looking at the wrong things.
• 29:42
  I mean, you and I both know this, but they're looking at the wrong way.
• 29:44
  Right?
• 29:44
  The benefit of having Copilot on your enterprise platforms is that it's preloaded with all the context because it's got access to all the information on your enterprise platforms.
• 29:53
  It's not that it suddenly is genius employee that can give you all the answers that you haven't thought of.
• 29:58
  Yeah, yeah, absolutely.
• 30:00
  And you know, and even then, the value of it is heavily down to the approach that the vendor has taken to pulling in and comprehending that information.
• 30:09
  So, you know, Microsoft Copilot, for example, absolutely will not be drawing upon the entire content of everything that's in your sharepoints every time you put a query in.
• 30:17
  That would be prohibitively expensive.
• 30:20
  And I think that's a thing that is quite invisible to a lot of people using it because it's definitely not part of the vendor's lexicon in their sales process.
• 30:28
  You know, ask a question of a small, semi randomly selected slice of your documentation, it doesn't sound good.
• 30:35
  So the framing is important.
• 30:37
  And I think the reason ultimately why PE and VC investors are unlikely to be fully automating their due diligence process, you know, in say, sort of two or three years time, is because how you package up the context
• 30:54
  for an LLM is so fundamental to the value that you get out at the other end, understanding how to ask the right questions sitting down and actually just Putting the legwork in, of talking to the people is such
• 31:10
  a crucial and totally unavoidable step in doing really good due diligence.
• 31:15
  I mean, our processes tend to run in a fairly cookie cutter way.
• 31:19
  You know, we will, every project is different, obviously.
• 31:22
  And so the questions that we'll ask are always very different, depending on the company.
• 31:25
  But we always start with a briefing from the team.
• 31:28
  We always get a demonstration of the product or services or you know, walkthrough of their portfolio of consulting or whatever.
• 31:34
  It is that the top, the one page answer ideally of how does your company make money?
• 31:38
  Right, that everyone should have.
• 31:39
  Absolutely.
• 31:40
  And then, you know, and then show me.
• 31:42
  But then we, you know, we, we look at the people that we've spoken to and we take a judgment on who, you know, who is really a decision maker.
• 31:49
  You know, how much influence does this person or that person have over the value that's being delivered within the company?
• 31:54
  And it can vary wildly.
• 31:55
  And it is definitely not a thing that an LLM can assess because, you know, there's no reason why an LLM would be able to, you know, there is no statistical vector graph waiting in the direction of looking at
• 32:09
  some conversation with a senior leadership team and assessing which one has the most decision making power over product direction.
• 32:15
  That's not a thing that is in, that is in the model, but it is a thing that's in our experience.
• 32:20
  And you know, we've been in plenty of situations where we've discovered that someone who is buried mid tier in an org chart is actually the pivotal person who is essentially critical to delivering all the things together for a
• 32:34
  team.
• 32:35
  So, you know, you feed all of the discovery documentation that you're providing in a data room into an LLM and it's going to do a pretty good job of digesting it.
• 32:43
  It will tell you what it's found.
• 32:45
  It can probably also do a statistically weighted, you know, output of telling you where the most likely risk factors are going to be on the basis of the entire corpus of its training material and other similar documents that
• 33:00
  then have some kind of corroborating output, you know, maybe a case study on the vendor's website or something.
• 33:05
  And if you have enough of them waiting, what it can't do is have eight hours worth of conversations with the delivery team directed by experience, drilling down into critical areas, making professional judgments as to whether or not the
• 33:17
  personalities and the people and their backgrounds and their skills involved are going to be the right people in 18 months time, and then drilling down into recommendations about how you assess that, what the milestone should be, risk signs
• 33:28
  and rag statuses against all of those areas and how it's going to have a knock on effect for exactly the same reason fundamentally as an LLM can't put together a financial model for you, it can write a piece
• 33:38
  of software that create an Excel spreadsheet that will be a good financial model because that's engineering.
• 33:45
  But what it can't do is it cannot get a contextual understanding and assessment and semantic analysis of the numbers and products and value delivery in companies is very similar to that.
• 33:57
  We use it all over the place for the areas where it is powerful and it allows us to increase our bandwidth and throughput and get our clients their results quicker.
• 34:05
  But we have not yet found any realistic way that it can replace a human's assessment.
• 34:09
  Yeah, and what's missed, the other thing I'd say is good at is what's missing.
• 34:13
  So you know, you have Billi and I want to talk more to you about Billi in a second.
• 34:18
  In terms of what Billy does, do you know, I have answers to it, my side as well, where I do use it, where it is valuable, you know, for example, if I've got a huge sprawling CRM database, then
• 34:28
  I can now run very reliably expansive data quality checks on like column completion percentage, whether something's used consistently, what type of thing it is, what's missing.
• 34:39
  And as you say, if you point at a data room like it knows what typically is in the data room, so it would be good at saying, hey, we normally see an org chart and we haven't got an
• 34:47
  org chart.
• 34:48
  So we're good at that, but that's the trap.
• 34:51
  And you gave a good example of a place that could be a trap.
• 34:54
  And you said, you know, an LM's not going to produce a good financial model for you, it will still produce a financial model for you.
• 34:59
  It might not be a good one, but if you ask it, it will produce one for you for sure.
• 35:03
  Loaded with all its assumptions in all its context.
• 35:06
  So that's the trap that people fall into.
• 35:08
  It is.
• 35:08
  And we encourage our customers, you know, customers and partners and target companies who say, yeah, you know, we're thinking about AI, what do you think we should be doing?
• 35:16
  Because part of the DD process is quite consultative.
• 35:19
  Right.
• 35:19
  You know, I think you go into the room, you talk to these people, you get to know them.
• 35:22
  It's hard to do the job unless you establish a rapport.
• 35:25
  And besides which we, we like people, we like talking to people about their work.
• 35:29
  And so naturally they will ask us questions about how other people similar to them and doing similar things.
• 35:35
  Our standard answer at this point around LLMs and how I think of it as a resource is that it is the enthusiastic, smart, but deeply inexperienced and slightly hapless intern sitting in the corner of the office that you
• 35:49
  can give, you know, if you give them very specific instructions they will take will go off and they will do that thing.
• 35:56
  If you give them a pile of 3,000 documents to digest and pull out the key points, they will do it.
• 36:02
  Occasionally they'll get distracted and halfway through the document they will absent mindedly start talking about the plot of the TV show that they watched last night and then get back on track.
• 36:11
  And you've got to be careful for.
• 36:13
  That although one sentence and decide it's the single most important sentence in the whole document when in reality is irrelevant.
• 36:20
  Absolutely, yeah.
• 36:21
  And it is essentially exactly what you would expect to do if you took a 22 year old business school graduate and you said, okay, your training for this job is Stack Overflow and Reddit and also Facebook and Instagram
• 36:37
  and TikTok and transcripts of everything that have ever been published to YouTube.
• 36:40
  You'd get a lot of useful patterns and correlation.
• 36:43
  And if you pointed in the right direction and said okay, this is the subset of that data set data that you should be using for context for this.
• 36:50
  But if you said to them, hey can you got a business that I want to sell some books, tell me what to do, then you're not going to get a business plan that's workable off the back.
• 37:02
  No, no, no, you're going to get.
• 37:03
  The zeitgeist of the entirety of humanity, which is not going to be particularly helpful to your use case.
• 37:09
  Yeah, here's the average best way to sell a book, not relative to you.
• 37:13
  So I said I'd come back to Billy, which is a kind of named product in terms of your example of the bits that can be systematized.
• 37:20
  Can you talk to me more about Billy and where you use Billy?
• 37:23
  Yeah, absolutely.
• 37:24
  So about 3ish years ago, somewhere thereabouts, we started thinking about how to scale our service offering.
• 37:33
  And as it turns out we're scaling in two different ways.
• 37:35
  One of those I think we've very briefly talked about is diligency, where we're building a consortium with our peers and that's essentially multi sector, multi focused due diligence and strategic analysis and transformation and so forth.
• 37:49
  I think that's a no brainer, not particularly innovative, just an avenue that we're using because we feel that by combining those disciplines together then we can give our customers a really strong joined up product that helps them make
• 38:03
  good decisions.
• 38:05
  The other avenue that we started exploring was the extent to which we could use LLMs, which were not quite immature at that point, but definitely emerging to increase our throughput and get us to the point where we had
• 38:21
  valuable investment committee relevant guidance and decisions and recommendations.
• 38:25
  Also of course, red flags quicker because I think anybody who works in due diligence or transformation or consulting and analysis will tell you that.
• 38:34
  Sure, there must be some people out there who have report writing as their favorite bit of the project.
• 38:40
  I haven't met them maybe, but it's.
• 38:43
  Introduce me if you find one.
• 38:45
  It's also, well, quite.
• 38:46
  I mean, I think if I found one I'd hire them.
• 38:50
  It's also immensely time consuming.
• 38:51
  Just sort of trawling through that information is, you know, representing the fact base from which you're drawing your conclusions and coming up to coming up with recommendations is.
• 39:00
  It's a lot of work, but it's also not necessarily particularly expert work.
• 39:04
  It's mostly consolidation and summarization, which is LLM sweet spot.
• 39:10
  So we started thinking about how we could use that internally.
• 39:13
  And for about 12 months or so before we started building Bilioneris, we ran that as an internal project where the document digestion and reporting and analysis phases of our work, mostly based on things like copious notes from discovery
• 39:31
  meetings and data and documentation and code based analyses, but also a lot of internal debriefs that we recorded and transcribed became the first draft for our reports.
• 39:42
  So we have a framework that we use internal if you like.
• 39:44
  It's the proto Billy that takes all of our analysis and team conversations and collaborative workflows and conversations about conclusions and recommendations.
• 39:53
  And it produces viable first drafts which we then do a lot of copy editing on and prepare and get ready to deliver to our customers.
• 40:01
  That's been a huge help.
• 40:02
  It's increased our throughput.
• 40:03
  It's meant that we can actually spend a little bit more time on projects really digging down into the detail and talking to teams and examining the assumptions which, you know, under other circumstances probably would be out of scope
• 40:14
  for a short DD project where we don't want to block the investment because there's a due date for completion, we can get more done.
• 40:21
  But the obvious thing off the back of that was, well, can we extend that?
• 40:25
  If we've got a system that has a really structured definition of what best practice looks like in areas where you can link that back to some definitive facts about a business, even if those facts change over time, then
• 40:40
  at any given point in time, if you've got the information, then you should be able to come up with a fairly straightforward set of metrics and guidance on how is the company doing on these important measures.
• 40:52
  Have they got the policies they need for this particular state of maturity?
• 40:57
  Does the org chart look functional?
• 40:59
  Are they about to?
• 41:00
  At least on paper?
• 41:01
  Yeah, absolutely.
• 41:03
  And the if you have conversational interfaces and automatic document ingestion and the ability to sort of give people little unobtrusive nudges from time to time, just feed from the channels that they're already using to provide people like board
• 41:19
  members with crucial documents about updates and finances.
• 41:22
  What can we glean from that?
• 41:24
  What are the facts that we can lift out and turn that into an ongoing, evolving picture of things like high level governance and operational health in that business?
• 41:36
  And is that a thing that's going to be useful to anybody?
• 41:38
  And through that research phase where we were using the prototype platform internally to do document reporting, we found that, yeah, there is, there's an awful lot that you can do around that.
• 41:47
  You can pull large amounts of information about a company's operational status and you know how mature they are.
• 41:54
  And I think very crucial things to investors like for the size and revenue this company has, how, you know, to what extent do they have their operational processes and governance together?
• 42:04
  If we invest and we hire a bunch more people and take them to the next growth level in the run up to a Series B, do we have an enormous maturity gap around just basic stuff like order reporting
• 42:15
  and decision making and sensible technical architecture?
• 42:19
  And does someone own the responsibility of figuring out at what point you need ISO 27001 or PCI DSS to keep on taking payments from consumers?
• 42:31
  You know, the stuff that seems really simple if you know it.
• 42:34
  And we do because we've done hundreds of projects already.
• 42:36
  And you know, even before that, you know, all of us had worked on lots of enterprise projects in sort of tech and systems before.
• 42:42
  But if you haven't done it before, if you're a generalist salesperson who's founded a company and found some techies and you're building something, there's no reason why you would necessarily know all of the moving parts.
• 42:52
  You don't know what you don't know.
• 42:54
  And the traditional way to deal with that is that you take on some money and then you very rapidly find a COO or someone very much like it, chief of staff, maybe or you find some consultants or you
• 43:03
  make a lot of mistakes and you hope that your investors are highly tolerant and will point you in the right direction.
• 43:08
  But if something comes up, like you have a series of board meetings with a critical gap and for some reason finance monopolizes the whole board meeting as is, you know, very often the case,.
• 43:18
  I mean, then PE boards are very finance background heavy as well.
• 43:22
  So yeah, I've seen the same thing.
• 43:23
  They look at the areas they're comfortable with and you know, everybody's got a boss, everybody wants to demonstrate they're adding value, even, even PE investors.
• 43:31
  And so things like how well equipped is this company to be able to actually deliver the thing that we've just plugged a bunch of money into, often that visibility gets lost.
• 43:42
  And so by building out a platform that knows the right questions to ask, that knows what the answers imply, can turn that into time series and risk statuses and tie that together into other dependencies and say, well, you
• 43:55
  know, if you've got this, but you haven't got that and you haven't got that, then in six months time when you reach this threshold for taking consumer credit card payments and you need to do PCI dss, you are
• 44:05
  not going to be able to do that in a period of time that will avoid you getting in trouble with the bank and possibly having your main revenue channel switched off for risk, for risk reasons.
• 44:16
  So, and there are thousands of questions like that in every business and some of those are not very important, some of those are very important.
• 44:25
  But if we can collect that information and nudge the right people at the right time based on our understanding of their org chart, then we can take stuff off CEOs plates.
• 44:35
  We can give a clearer picture to boards without the CEO having to spend loads of time collating nitpicky details.
• 44:40
  Together we can track trends and changes and org chart health and gaps in responsibility over time to look at the maturity of a company and whether they're getting to the point where they need to start thinking about something
• 44:51
  with a little bit more importance and we can help them build capability as well.
• 44:55
  And I think we were asked a question when we were on a research call a couple of days ago by a startup CEO that we know quite well and we were explaining what the plan was and she said,
• 45:06
  well, this sounds very much like you're just sort of offloading capabilities into a separate system.
• 45:12
  It's just another thing.
• 45:14
  I've already got loads of things that say we'll take this off your hands, but that's no use to me, because I need to be able to do this.
• 45:20
  I need to know that we've got this expertise in the company.
• 45:23
  And our answer is pretty straightforward, which is that that was very obvious to us from the get ago when we started building out the company.
• 45:29
  And the answer is that, you know, it's a system which focuses not on doing things for you.
• 45:35
  There are plenty of those around.
• 45:36
  Anyone can go and Sign up for Monday.com yes they can.
• 45:39
  And you know, spend a whole load of time building custom integrations or getting a integrator to build a whole bunch of automation workflows.
• 45:47
  But that removes things from the responsibility of your team, but also makes it largely invisible.
• 45:51
  It falls over, then you know it's out of your hands.
• 45:55
  Billy is specifically designed to spot where you've got gaps and then support you in building around those gaps so that your company gains the capability and closes that gap.
• 46:05
  Because then your people are driving your business forward.
• 46:08
  Your technology is the thing sitting in the background acting almost as a virtual chief operating officer, going, oh, you know, we've got a skills gap over here, we need to fill that, guiding them through the process of doing
• 46:18
  that.
• 46:18
  And all of that can then be fed back to portfolio owners so that they can see where they need to step in and offer support, rather than saying, well, we'll just solve all of your problems for you for
• 46:29
  a very reasonable monthly fee of $12,000 a month for your team of 30 people, which is almost inevitably where all of these kinds of solutions end up going.
• 46:39
  And CEOs have had their fingers burnt before.
• 46:42
  They've taken on Intercom for their customer service conversations because it's affordable and it's flexible.
• 46:48
  And then they get up to the point where they have have five customer service representatives and they find that they're paying one and a half thousand dollars a month for a fairly straightforward set of functionality.
• 46:59
  We believe that we're building a much more sustainable business and a much more attractive one for people engaging with it and using it to build capability within their businesses and also their investors getting a good view of a
• 47:09
  portfolio because the focus is on enabling teams to fish themselves.
• 47:17
  We want to give people the fishing rod and the skills and the ability to carry these responsibilities policies forward and then keep an eye on that over time, rather than just taking it off their hands and leaving them
• 47:29
  fundamentally less agile, less capable, less skilled in exactly,.
• 47:34
  Exactly the same way I set my business up.
• 47:35
  And in fact I'm open about it of like, the last thing I want is to build a reliance on my team.
• 47:39
  Right.
• 47:40
  There's an exit plan here for all of us.
• 47:43
  Yeah.
• 47:43
  And I think, you know, there are a few good reasons for that as well.
• 47:45
  I mean, because we've seen so many companies go through these growth phases.
• 47:49
  Every now and then you get a very high profile success story like a Monday.com or an Atlassian or Salesforce where they've been very successful at being all things to all people and arguably they've become so flexible that actually
• 48:04
  they're a little bit inaccessible to the long tail of the mass market.
• 48:09
  They're actually now really complex and you sort of need an integrations partner.
• 48:12
  So they're missing out on quite a lot of revenue from a practical perspective.
• 48:15
  But also that you know, they are creating lock in and there is only so long that you can go as the provider who's locked everybody in before.
• 48:26
  Eventually they get big enough and capable enough that they get annoyed and either move vendor or they develop you out of a contract.
• 48:34
  Whereas, you know, we've seen people go through these phases.
• 48:37
  Those success stories are relatively rare, which I think is why we hear about them because they are exceptional.
• 48:43
  And actually much more common is companies who take exactly the same strategy because they see those vendors doing that as a SaaS play.
• 48:49
  And actually what those people find is that everybody looks at their business model and goes well no, I know how this goes.
• 48:54
  I'm not doing that.
• 48:55
  I want a platform that will support me and grow with me.
• 48:57
  I don't want a platform that's going to lock me into their way of doing things forever and force me to have this massive bottom line cost forever.
• 49:05
  And so we would love to be a salesforce in terms of size.
• 49:10
  You know, we want to be that successful obviously.
• 49:13
  But you know, we think that they're much more dependable and also I think the much more defensible and this, you know, correct approach to this is to build a platform which is a collaborative partner and works with customers
• 49:25
  and doesn't lock them in but instead delivers the value that causes them to stay with us because it's good, because it helps and because it fills a gap.
• 49:32
  Not a fan of software.
• 49:33
  That is basically mask is software masquerading as consultancy because in order for the software to work I have to come in and change all your internal systems in order to get it to do its job.
• 49:43
  Exactly.
• 49:43
  So I love that people want to find out more about Billy or the diligency group or you indeed rich in your path from audio engineer to private equity tech due diligence AR guru where.
• 49:57
  Should they go look lots of websites.
• 50:01
  I am rich kershaw.com and diligency is diligency.org we're also online@lecky kershaw.com which is the tech ops and CS.
• 50:13
  You can also find me on LinkedIn as well.
• 50:15
  Billy is currently in the build stage.
• 50:18
  We're hoping to go out to a beta audience in about eight to 12 weeks somewhere thereabouts.
• 50:24
  So if this sounds exciting to you and you would like to be part of that, then I would.
• 50:28
  You know, we would love to get enthusiastic, practical, pragmatic people involved in the, in the beta test program because fundamentally our big priority is getting a.
• 50:39
  Getting a software platform up and running that actually genuinely really helps startups, investors, you know, One man band enterprises, all the way up to companies going sort of through to their series B.
• 50:51
  And if we're solving real problems that people really have and filling those gaps and reducing the overwhelm for founders, then we're doing something right.
• 50:58
  And a big part of that is making sure that we get the right feedback once we launch and go to market.
• 51:04
  And you can find more info about that at Billi Works.
• 51:06
  That's B I L I works.
• 51:09
  Nice.
• 51:10
  Thank you very much and I appreciate you coming on the show and sharing such a great wealth of knowledge and insight for everyone listening.
• 51:16
  So thanks very much, Rich and I know we will talk again soon.
• 51:20
  Pleasure.
• 51:20
  All right, thanks so much, Gary.
• 51:21
  Take care.
• 51:22
  Bye.
• 51:23
  Thanks for listening to the PE Data guy.
• 51:25
  The place where private equity meets data.
• 51:28
  Please forward this episode to your favorite private equity friend.
• 51:31
  Thanks for listening.
• 51:32
  See you next time.