I asked ChatGPT to generate a professional athlete's statement after losing a big game, and this is what I got back:
I told ChatGPT to modify the statement, making liberal use of the words “kinda” and “obviously,” and I think that we can all agree that it looks more realistic ;)
[Jim Milbery]
We are speaking with Alan Williamson about his new book, "How To Think Like A CTO." What was the impetus for you writing this book?
[Alan Williamson]
I've worked with you for nearly 25 years, and you've taught me a lot, and when we go into portfolios and look at companies, we see the same problems repeatedly. This is a book on how to run a technical organization and where all the issues we've identified come from. Coming from this world myself, I wish I'd had this book ten years ago when I first became a CTO, particularly for a PE firm, because of the questions being asked of me, and I had no clue how to answer them. When I started, I had no clue what a CTO does.
[Jim]
I see CTOs as visionaries and architects, but maybe not the best people managers. They might work well with small, hand-picked teams, but they're not about keeping the trains running on time, detailed project management, hiring and firing people. The VP of engineering tends to be the person that owns the project plans on track. So we see those as two different roles, but they're often a single role, given the company size. So what is your definition of a CTO?
[Alan]
I would say for companies of the size we deal with, those terms are usually interchangeable depending on the flavor of titles they use within the company. But a CTO is there to determine the "how" of the journey. And I would say that a CTO at this level might be running an engineering team, but they also have a VP of engineering for the day-to-day tasks. But fundamentally, they're there to ensure the technology is going to a place where they won't get themselves in trouble three to four years from now.
[Jim]
CTO, it's a title that's thrown around pretty liberally. When do they get hired into a business, and what is a person's path to becoming CTO?
[Alan]
It depends. We've seen different flavors of it where you have someone we'll call Billy who has never left the company in 15 years. They're the most senior person simply by staying there. He's built up a team around him. Still, people of Billy's type will never hire somebody more intelligent than they are, so if Billy hasn't introduced this technology, framework, and language, it will not come into the company. At that point, yes, you need to bring a different flavor into the cake to mix it up.
[Jim]
So let's highlight that in terms of your book. Did you think about it for the poor bastard who is replacing Billy?
[Alan]
Yes. There's a whole chapter on taking over from an existing CTO to gather all that knowledge. And again, it's about celebrating the fact that they built something. There's no point in criticizing any prior effort whatsoever. They've got to where they've got to be, so celebrate that. I also delve into situations where the CTO is the wrong fit. For example, suppose you're more of an engineering-led organization with a CTO that is primarily a consumer of technology, such as third-party services. In that case, that's not a good fit to lead an engineering and product-led group.
[Jim]
How does somebody figure out that they are more of a "consumer of other people's services CTO" than a "building it from the ground up" CTO? Especially in a world with so many open-source solutions and libraries?
[Alan]
I also go through that as a chapter, which is the buy versus build mentality. I've included an assessment to help determine where you live in that world. We can get seduced by looking at a framework and saying, "Oh, that's going to do everything." But we spend more time trying to get it to do what we need to do rather than building something new in the first place. So there's always that pull and push to figure out what's what. But fundamentally, it's what's in your DNA. If you're somebody that builds stuff, you will naturally gravitate toward that type of solution. But we must temper that by saying, "Do I need to build everything myself? Where can I get major gains because I no longer need to build that component?" We've seen the evolution now where the amount of code you need to do in an AWS and Azure environment has collapsed dramatically compared to 10 years ago when you would have had to manage much of this stuff yourself. To be able to have the line of sight to be able to figure out where and when do I buy, do I build, do I outsource, do I consume? That's a complicated question these days.
[Jim]
So let's also flip that on its head a little bit, you're trying to make that decision yourself, build versus buy, but as a CTO, you're also trying to explain it to a board or a C-suite that may not understand any of the ins and outs, the tradeoffs, and the risks. How does a CTO get to that language translation?
[Alan]
Most technologists fail in this situation - they don't know how to talk to the business without using buzzwords. And I go through that in the book in extreme depth. I use a lot of metaphors instead of technical buzzwords when I'm explaining something. And it's crucial that the business can feel like they've got a say in the decision. You never want to go to the board and say left or right. You must give them the knobs and sliders to make them feel part of the decision. Get them into it; half of that is laying out the options in non-technical terms and with a firm budget. Presenting senior management with three simple facts is the key:
This is what it's going to cost to build it.
This is the timeline.
This will be the cost to keep maintaining it from there on in.
A part of the process many people forget is after you've built it, then what? You have to support the bloody thing and ensure you continue to hire for it, for when the time comes, and you move on, someone will still be able to maintain it.
[Jim]
Let's return to your original point for a second. You're giving the business three primary weapons. First, you provide the business choices and explain the tradeoffs, but then as CTO, you've got to live with that choice. Don't give them options if there's only one you're willing to live with. You need to avoid using buzzwords. And if you use metaphors to help with the translation, you better be good at it. Almost nothing is worse than getting dragged into a pointless discussion from a bad analogy.
[Alan]
And funny enough, I have a chapter about communicating your vision and your complete vision statement and getting that rehearsed in your head. Make sure your vision statement for the company is that elevator pitch that you can say to any employee in the company, and they'll understand where you're going. Most of the time, it's what we can deliver and what we will improve. You're not promising that everything will be perfect in version 2.0. That is setting yourself up for failure right from the start. But it's honing that message so that it becomes much more natural when you're saying it, and therefore the metaphor should become less contrived and much more natural when presenting it to non-technical people.
[Jim]
So the last piece of this is that you, the CTO, are making long-term decisions for the company that might outlast your time as CTO. That's a tricky thing for a CTO to do well.
[Alan]
Yeah. And because most CTOs think in project terms, that's the big thing. Give me your vision of where our platform will be in five years; most likely, we'll struggle with it. And that's always my litmus test when talking to a CTO. Can you hit the core of the problem if I was to give you a whiteboard? Could you assess the company now and where it needs to go? And that's where most people fail. They'll get into the minutiae and explain to you what they're doing in the next year, but that's not a vision. You're thinking more like a project manager than an architect.
[Jim]
LinkedIn is the only social media site that I use now. In the past, I've used Facebook, Instagram, and Twitter. I dropped off Facebook and Instagram back in 2018. They were great for sharing things with family and a small group of friends, but the benefits weren't worth the costs to me of being a Zuckerberg customer. I was a Twitter lurker for many years -- read many posts but did not post all that often myself. I closed my account when Elon Musk acquired the company because, well, F-T-G.
I created my LinkedIn account in 2003, so I've been a user for twenty years. At first, I tried to carefully curate my account -- only accepting connections from people I knew. My viewpoint back then was that anyone in my connection list was only a phone call or email away. If I reached out, they would respond. Ultimately, it was an impossible task. To wit. I am on the operating side of ParkerGale, but I still get inbound connection requests from bankers. There is no upside in turning down their connection request. At first, I'd steer them toward one of my deal partners in a polite email response. That was too much work, so I started accepting more requests than I wanted. I still turn down more requests than I take, but my connection group is still far less curated than I would like. This has the side effect of making my feed longer.
I've been a premium member since 2016. Premium gives me InMail access and the ability to see who has viewed my profile. I'm not sure that's worth $59.95 a month, but it is what it is. Despite being a paying member, LinkedIn fills my feed with a metric ton of advertisements. At this point, every third or fourth post is an advertisement -- and almost all of these are auto-play videos. I disabled auto-play, but video images take up valuable screen real estate. It's a real teeth grinder.
You can hide any advertisement, but LinkedIn only allows for two reasons for hiding an ad:
This isn’t what I want; I want more central control over my feed. What I want to be able to do is block ads at the COMPANY level.
Let me give you an example. I hate Pepsi and don’t eat salty snacks like potato chips, so I don’t need to hear from Pepsi’s Renfield (Frito-Lay), either. When I ask for a Diet Coke at a restaurant, and the waitress responds by asking whether Pepsi is ok — I say, “No, it isn’t ok; I’ll go with ice water.” I hate Pepsi products THAT much. If LinkedIn shows me a Pepsi advertisement, I want a radio-button choice on the “hide” panel that says:
“Do not show me ANY advertisements from this company”
And then, I don’t want to see Pepsi ads on my feed ever again. Full stop.
It’s good for Pepsi AND me. They don’t need to waste resources showing me advertisements I will never click through. Sure, Pepsi might come out with a future product I would love, and I’ll miss out on it. I’ll take that chance.
I’m not an ad-tech guru, so I am guessing there might be challenges in making this all work. But this is what I want. It is much easier in the flow of reading my feed to simply “block” companies I don’t want to hear from than trying to fill out a profile of companies or product categories I want to see.
Maybe LinkedIn uses machine learning and AI to de-emphasize companies that show me “annoying or not interesting” ads. I want to be explicit about it.
[Jim Milbery]
I am here with Lukas and Dan from Code & Co. We've worked with you on several projects, and we're about tech due diligence today as part of a broader overall diligence as you look at companies. Lukas and Dan specialize in tech diligence, so I thought they were the perfect guests to discuss the ins and outs.
Maybe you can introduce yourselves, talk about Code & Co, and then discuss how you chose to focus on tech due diligence for private equity guys.
[Dan]
My name is Dan, and I am one of two founders of Code & Co. We founded the company seven years ago, and we specialize in tech and product due diligence, and even more so, we focus on private equity and growth funds. And thanks to our industry, we can do so remotely. So we're a remote-only organization, and we work with funds worldwide.
[Lukas]
My name is Lukas, and I am a cofounder of Code & Co. I worked as a software engineer at ThoughtWorks before founding Code & Co.
[Jim]
How did you both meet?
[Dan]
We met around 2010 or so. Lukas and I were at the same VC office begging for money because we both had founded a company and were raising funds. I did something in digital publishing, Lukas did something in e-commerce, and this is how we met. What's funny is that we clicked personally, but professionally, we went our separate ways. So I did an executive MBA at UC Berkeley and the Technical University of Munich and then joined Rocket Internet in Berlin as a director of products. I built several companies from scratch when Lukas was a software engineer at ThoughtWorks. We continued talking and always figured we wanted to do something together. In 2016, we discussed building something around understanding software and building technology; hence, Code & Co was born.
And I think the story about how we came into doing tech diligence is also quite interesting. We played with a bunch of software solutions, but then a friend of ours, an M&A adviser, reached out while raising funds for a company. And he didn't understand what it did. He said, "Guys, you know technology and business, so help me translate." We didn't know to call it due diligence, and since then, we have used our network to grow the organization.
[Jim]
So let's talk a little bit about your methodology. We buy founder-owned software businesses. So our focus is on the software as the product. Before ParkerGale, Devin and I were in other funds that did other deals besides software, such as healthcare and manufacturing businesses. We also did technical diligence on those deals, frequently leveraging packaged software, so we had very different requirements. For example, we owned a company that made metal roofing products, and they were on Epicor as their ERP platform. So, they didn't do any custom software development in that case. We focused on finding guys who knew Epicor to ensure a proper implementation. You primarily focus on software companies instead of technology at non-software companies. Would you agree?
[Dan]
Absolutely. We focus on software and do everything from deep tech, AI, ML, big data, cybersecurity, and API life cycle management, to tech-enabled offerings such as IT services. Ultimately, we always say we look at companies where tech is the product or could be the product in the future or where tech plays an integral part in accelerating the value. In the case of an IT-enabled service, they have internal software or they have accelerators they can use. So quite sector-agnostic, but yes, the software is the key.
[Jim]
If you're a manufacturer running a commercial ERP system like SAP or Oracle, I would still do InfoSec diligence and focus primarily on finding somebody who knows that ERP system to audit. I would say the tip of the spear at this point is, no matter what, you should be doing InfoSec diligence. Is that fair?
[Dan]
Yes, it's fair. In addition to that, we're not big fans of something that people could call IT due diligence or pure digital due diligence. In the end, humans build software, and teams scale software. And then users use software because it alleviates some pain points, right? And so we don't want to do just an IT due diligence that assesses a company's phone systems and file servers. We're trying to optimize for impact and look at the opportunity that the company is pursuing. Tech plays an essential part in value creation in most deals in product due diligence these days. Hence, we believe it must be a holistic, modern, and actual assessment comprising technology, product management, and engineering. We define product management as the art of knowing what not to build. So this means can a company allocate its resources effectively? Do they make something worthy of the customer's attention? Does it alleviate a pain point? What interfaces do exist? Do they have the data to back it up? What's the plan going forward, product vision, and road map? And what's the customer onboarding process like? And tech - very much the hard facts from architecture to infrastructure, automation and monitoring, scalability, and single points of failure. You mentioned security, and we look at that, too, of course. And lastly, engineering, which is people and processes. Who's missing? Or maybe who should not be lost by the company? Who should not leave the organization? Software development has a life cycle so the keyword would be agility and fast iterations.
[Jim]
Most employees will not know there's a sale process going on officially. The chatter spreads typically the closer you get to finishing a deal. But in reality, we're trying not to engage with everybody in the firm. Because that's true, there are just some things we won't spend time on. You focus on critical systems. So as I divide these things into pieces-- if you're buying a healthcare company, you will focus on their EMR and EHR systems, and you're going to get experts who know those systems. If you buy a manufacturer, you concentrate on their ERP and supply chain software.
[Dan]
[Jim Milbery]
So, Matt, you have a great background; give us a quick commercial on yourself and DCP so the listeners know who we're talking to today.
[Matt Lynch]
I started as an investment banker at Solomon for about 11 years and then had an opportunity to run corporate development at a tech company in Washington, DC, called Blackboard. I was there for a few years, and that's when the lightbulb went off for DCP. DCP is a merchant bank, meaning we have an advisory and investment arm and are almost entirely buy-side-focused. We take an in-house perspective to everything we do, which we think is a differentiated niche. When considering starting DCP, I looked for a similar model; there's not much out there. There are many brokers and outreach firms, but no one we've found who manages the entire buy-side process from start to finish. We work with clients on a programmatic level, meaning we can drop in a team and support, build, and manage a playbook. Or we can work on an episodic basis, i.e., execute a single deal, as you would typically expect an adviser to do. We manage a portfolio of investments based on DCP advisory deals, and I think it reinforces the empirical nature of what we do. We're not trying to get any deal done for the sake of doing a deal. We want to get the right deal done, and often, we have our skin in the game to emphasize that point.
[Jim Milbery]
So today, we will talk about roll-ups. First and foremost, let's define what an M&A roll-up strategy is. Define roll-up for us from your worldview.
[Matt Lynch]
An M&A roll-up is a strategy for acquiring a platform company. Then that platform will often acquire numerous smaller companies, which drives a lot of value creation in the form of financial returns. It sounds simple, but there are multiple paths to do this. First, the platform is typically the foundational acquisition. But you can put that together either with a significant initial investment, or you could put a couple of smaller subscale companies together and Frankenstein the platform into being and then go off and find add-on acquisitions.
[Jim Milbery]
So, what's the platform deal's minimum size?
[Matt Lynch]
It varies by market size and fund size. Different private equity firms are going to go after different market niches. But I would say, on a relative basis, the add-on acquisitions are probably somewhere in the range of 5% to 40% of the enterprise value of the platform. If you think of the platform as the basket you're putting all your eggs, that should give you a sense of size.
[Jim Milbery]
Now, let's talk about different roll-ups. So when you look at this, are you buying features, taking out smaller competitors, or buying market share? How do you see it from a market standpoint? And does that vary between an industrial company and a tech company?
[Matt Lynch]
I would probably put them in the buckets you described. I'd say there's the acquisition of direct competitors. So you're looking to acquire market share in your existing market. You're adding customers. You're taking out direct competitors like JetBlue's pursuit of Spirit Airlines or the T-Mobile Sprint deal, where you're directly complementary. And then the second bucket is a vertical acquisition—something where you're building or adding on features that are missing in your product. When eBay bought PayPal, that was a deal where they debated making their payment mechanism and decided instead to buy it. That would be a vertical expansion or add-on acquisition. And then another bucket is adjacencies, which is extending your product into new markets or geographies. What was your customer doing five minutes before or five minutes after using your product? And buy that. Facebook bought a company called WhatsApp, which was outside its core business. They had a messenger but were trying to build adjacent market share, which is another good way to expand your total addressable market. And those are the three big ones.
[Jim Milbery]
Yes, and indeed, the economics will vary. I could buy some tiny competitor, but if that feature is a feature I could potentially sell to every one of my customers, I can make money on the arbitrage. If I think, "Wow, I can get 50% of my existing customers to buy this feature," then it's super simple. We had a portfolio company that did a ton of that. They bought features for a dollar, sold them to customers for two bucks a piece, and made solid margins. The customers got value from the new features bundled into the platform.
[Matt Lynch]
I think that's fair, and we did that at Blackboard too. One of our clients, Diligent, is a textbook study for this strategy. They started in a board portal space, which was highly fragmented. They bought up many of the other board portals, then expanded their category in the governance angle. And that's another example of fundamentally changing the aperture of your company. I firmly believe that 2023 will be the year of the add-on acquisition. I think we'll see a lot of carve-outs where you've got non-core assets to divest. As businesses reevaluate what they have and want to move forward, you will find many non-core assets they would much rather monetize in this market environment and get the capital to deploy elsewhere. So consolidators can take those non-core assets and either create them as add-on acquisitions or make them the platform. Another company we've worked with called Probation was a carve-out of a large healthcare conglomerate that wasn't interested in participating in software. It turned into a very successful platform for Clearlink.
[Jim Milbery]
I think you're correct; corporations who acquire a company that's acquired somebody who's acquired somebody, etc., now have an asset that's not core to their business. Now is the time to get rid of it. I also think we'll see a lot of walking dead on the venture side. They're not dead, but they won't get any more investment from their venture investors. Now it's time to decide what the best owner for them is. Customers like one throat to choke, and large companies like Oracle have successfully bought many features via acquisitions. Oracle customers would like to buy from Oracle. So maybe they bought a walking dead product from venture guys that wasn't the top product, but it was better for their customers because it is now tightly integrated with other Oracle products. So there are probably a lot of "feature" companies out there that will be for sale that wasn't going to be for sale in 2022.
[Matt Lynch]
I worked on the PeopleSoft defense, an old sale to Oracle. These are companies that aren't for sale. You look at Oracle alums and what they have done and what they have built. It's incredible how that organization has seeded a new generation of software behemoths.
[Jim Milbery]
The funny part is I worked for Ingres way back in the day and later worked for Oracle. Ingres, Informix, Sybase, and Oracle all hated each other back then. What's funny is we see the other side of it now. Many people who worked only for Oracle tried to come down to small Private Equity backed companies and struggled because they lacked the resources they had with Oracle.
[Matt Lynch]
And actually, that's another tie-in if you found a market niche that you think is reasonably attractive and ripe for this roll-up strategy. You have incremental resources, shared back offices, and higher margins for the most part. You can do the types of things that Oracle does on a much larger scale to push your competitors around.
[Jim Milbery]
As we start to talk about how you get started on an M&A strategy as an investor, should you have a hypothesis and market in mind before you even think about doing this? Or can I go to a provider like you to give me that hypothesis and get started?
[Matt Lynch]
Loading more posts…
