Duncan Davidson (Bullpen Capital): In Silicon Valley failure is a feature, not a bug.
04 Feb, 2022
Steve Berg is Partner and Managing Director at Lytical Ventures. Prior to joining Lytical Ventures, he was the General Partner of Antecedent Ventures and was part of the RTP Ventures investment team. He has held positions in strategy and corporate development at Emulex (NYSE: ELX) and as a sell-side analyst covering technology at Punk, Ziegel & Co.
I would see that there is no single glide path – there are so many different ways into venture capital. My career started as a sell-side analyst on Wall Street is covered storage in storage networking companies. I am CFA and MBA, so the time value of money doesn’t come hard to me at this point in my career. I then run corporate development for a public company for a number of years and started doing some mezzanine stage, or later stage investments, which is the right of first refusal for acquisition. And then I got to do a little bit earlier, earlier, earlier and I realized that my continuum starts at behavioral psychology and goes times value of money. Frankly, I liked the behavioral psychology part more – I like rolling up my sleeves, I like the operating part, I like the mental challenges of getting a startup off the ground so much that I created Intangibles podcast which is all about behavioral psychology traits that I think are important. My first investment fund was RTP Ventures, it was wildly successful, and the managing partner there was kind enough to let me start a seed-stage operation. Ultimately it ended up growing to be Lytical Ventures which is what I’m doing today.
I think everybody’s fear at the beginning is “Will we see enough deals? Will we see enough quality deals?” I can go into the nature of quality deals. Everybody thinks they know what are the quality deals until they realize that there are even higher quality deals behind the curtain. But I think what you realize relatively quickly is that if you’re legitimately writing checks if you’re legitimately bringing something to the table, enough deal flow will find you. I think I was initially concerned about that but not anymore.
I would say that the world I invest in is a world that I kind of invented. My background is data, analytics, some cybersecurity, and machine learning – that’s where I come from, that’s what I’ve been traditionally good at. Frankly, it’s all I really know how to do. In terms of the exotic nature, I just don’t venture outside what my area is. And I think what people are doing with machine learning these days is phenomenal, it’s world-changing, deflationary in its nature, in terms of productivity, in terms of seeing things that we have not been able to see before. I’ve seen some crazy amazing demos – even on the cybersecurity side! Like some of the cloak and dagger things that the US government is doing that makes my mind explode. So I would sadly say that I haven’t really invested in crazy or exotic things, but my normal world and companies there are super interesting.
Once you hang up your shingle, inbound is inevitable – folks looking for capital do pattern matching and trying to find us. We are internally a little group of folks, or band of merry men, that put our heads together and there’re definitely places that we want to invest. And we are really keen to make sure that we had a bed and third-party risk, and we are looking for that. For some particular things there we feel as though we can see what the future looks like, and we go after those things, trying to find them. But if you looked at every company in our portfolio, if I gave you the origin story of each of them, it’s a mix.
I really don’t even know, but we do keep records. We do keep all our deals tracked in an equivalent of a CRM, so we know how many came in, how many go into the top through our filters, how many times we say “Yes,” what’s the percent. The truth of the matter is that I don’t really care about how many times I say “No” related to how often I say “Yes.” Some people want to know, limited partners are often curious about it. But the ability to recognize something and say “Yes” to it, I think, is independent of that. I would almost argue that you’re not doing a very good job filtering if you’re looking at a whole bunch of things that you said “No” to. The last time we reported to our LPs was in usual for funds 1-2% range, but it’s not a metrics I care about or remember.
At the beginning of every fund I made the architecture of how much capital is going as an expense and how much can we actually invest. If you think about check size or ownership you like to target at every stage, it gives you how many deals are going to be in a fund. For our first fund, our expectation was that we would have 15 deals. We’re reaching the end of our first fund now. If you think about the investment period, that’s 3 years for us, it makes 5 deals per year. I suspect that our second fund will probably have the same architecture that has 5 deals per year. You need also think about sitting on board, really taking part in the company development, and how much can a partner handle relative to other things like maintaining good relationships with LPs, cultivating new LPs, finding deals to invest in. When you think about that balance there are probably 6-7 companies that can be active and helpful. So by the time we’re at the middle of fund #2, this might be full unless a few companies will exit or I step off their board.
We are looking for areas where large data sets are used – large data sets in Finances, Healthcare, commercial companies as well. So, the general answer is I’m interested in verticals that have the most data, where the most value can be derived from data analytics and from AI/ML.
I don’t think we have a strict “No,’ but I do know that the likelihood that I will do B2C technologies is quite low – that’s not where my skill set is. Probably not even B2B2C if you really want the truth. A lot of people, even those that do stuff in Cybersecurity, think they will never do things with the government. This is the huge market in the government spend and they did not stop spending at all recently, during the COVID, where some people may have cut back or slow down. It is a lower multiple on government revenue. I’m a little bit more open-minded and if I can see a commercial application of something that happens to have gotten a bunch of government revenue, I’m fine with that. If it is a company that works just on government contacts and that’s all, it’s not for me.
Fund #1 is an interesting animal: you need to make sure that you’ve got enough different companies and different possibilities to show that you can do what you say you’re going to do, also to have enough diversification to withstand companies that don’t work out. For our first fund ($35-36M) we decided not to follow on, to make initial investments in seed stage, Series A, and Series B, but any pro-rata rights we will pass to LPs. We make sure that we’ve got 15 deals in this portfolio and making sure that we got reasonable early ownership, that we wouldn’t muscle out of deals. For fund #2 we have made certain that will be able to follow on 1 round for sure and then, in certain instances, possibly 2 rounds.
It’s different at the stage that we invest at. For the seed stage, for example, there is less diligence to do on the financial side and on the customer side, but there’s a lot more diligence on the tech and team sides. When I think about the team I think about that behavioral psychology part: can you ascertain that this is a team that is great, that is going to get stronger when things get harder, are they open-minded and flexible and have the humility to listen to what’s going on in the environment and adjust to that. However, as the company matures and its processes mature, you can now start to look at other people in a company, you can start to look at contracts and what kind of contracts the company does and their quality, you can look at the growth rate, you can look at the pipeline, you can look at the way they track a pipeline and move deals through, and get a confidence level for their ability to predict their own feature, whether it’s 3 months or 3 years. The diligence process gets more thorough, there are more things to do. It usually more depends on the company how long does it take, how quickly does the company respond to our questions and document requests. Ideally, it’s complete prior to issuing a term sheet – some firms don’t work like that, but we prefer to. Valuation is understood before that as well. I think it could take a month, but it could be shorter, it could be longer. if it’s Series B and there is a repeatable sales process – it’s more to look at.
In fund #1 seed stage check was $500K, Series A – $2-2.5M, and Series B check was $3-3.5M. Those sizes will go up with fund #2, because we like to get more ownership, we actually get the more dry powder to put to work, and as deals get larger in size, a smaller check is just less competitive in the market.
We are a fund that is investing in pretty DeepTech, and most of our founders are technical founders, whip-smart, and they can build a thing and also have the vision to understand that other people are going to try and build that too and by the time they get to where you’re at now you need to be someplace better. I personally like to look for what I call “inside/out” – somebody on the inside buildings a tech and someone’s going on the outside dealing with the different constituencies, like dealing with customers, making new hires or investors, focusing on running the business. The team that has different skill sets has a greater chance of winning and being successful. I don’t think there have been many situations where we invested in someone who is just a business person who kind of contracted tech side somewhere else.
There’s a fine line there. If somebody adjusts to every piece of feedback, I don’t think that’s necessarily right. There are people that don’t take feedback very well at all. In our portfolio, we probably air on the folks that are going to live and die under their decision, and it doesn’t work out it’s on them. But I do think that even if when you make a suggestion it takes a while for them to fully ingest that suggestion, I do like people who are constantly analyzing their external environment to make sure that all of the assumptions that they made are still valid.
Folks that don’t have at least a decent idea of what you’re asking them about their business at the time you’re asking. The founder who doesn’t know what their revenue is for the quarter or for the last year seems weird to me. That’s your lifeblood and you don’t know the answer? You’re not fully in touch with the specifics of your business? I want someone who’s paying attention. I also want someone who’s got multiple points of view, who gets input from different sources in different viewpoints, who has a broader outlook. An all-one-point-of-view person makes me worry. My job is to minimize risk while maximizing return, and I’m looking for things that can increase risks – behavior-related things, operating, marketing things, etc. I see those things that are raising my risk while my return isn’t changing commensurately, I’m concerned.
All of the guys that work with us on our side of the table have been, at one time or another, an operator and can provide many different types of help to founders – if they need to find a product-market fit or hire an HR or with insurance for fundraising, etc., etc. And depending on what’s required our level of engagement changes. I’ve got that far as dropping into a company and working as a part of the operations there. I’ve only done that once, frankly. And at the same time, it may be as high level as looking at the deck every 6 or 12 weeks. Sometimes founders just don’t need our help, and we’re not in the position to push ourselves upon them if they are doing well.
The goal of an initial pitch is to get another meeting. And a lot of founders think that the goal of the initial pitch is to get the VCs to write a check. That just doesn’t happen. I think that kind of a little bit overzealous, overaggressive, maybe a little naive approach doesn’t win very much. My advice is to have a little more patience, revise the goal to continuing the conversation, to getting to know somebody a little bit better to have an opportunity to share more later when you can start to dig into the details and get to the things that make it interesting.
Actually, you’re asking “When should a fundraise commence?” The market for fundraising is very broad, and there are companies that can get this done in a couple of months while others consider it’s going to be months of talking before they will get a couple of term sheets, will go back and forth trying to figure it out, then finish diligence, and then there’s another month to get the paper. The company knows what kind of company it is. If you’re of the first type, you can probably run it a little closer to the edge. If you are that second type, you should be keeping 4-5 months of burn available to you.
Internally we push for a whole rate of about 46% IRR, which means 3x to 5x at 3 to 5 years. The rate of return that the market is comfortable accepting tends to be less, so 3x return on investment after a couple of years, we’ll be fine with that. I know there’s a lot of companies out there that want to have a billion-dollar exit. If you’re a $36M fund, like our fund #1, to get a multiple of return you don’t need that gigantic exit, but if you have a higher percentage of the ownership, you’re going to win a little bit more often. I would say that we’re not as aggressive in swinging for the fences as some, still, we really do shoot for above-market return.
At Don Valentine’s days, the golden days of VC, there was permissible to own 25% of a company, even more. I don’t think that’s the market anymore. If a founder sells 20% of their company in around A and then 20% in the next round, that seems closer to reasonable. There are things you can do to drive your percentage of ownership up, but you don’t want a founder who’s given up so much of a company that they’re not motivated, they don’t have the skin in the game that they should have. Back to red flags: if we see a founder that has sold too much, that’s a red flag.
It’s my first fund, so I didn’t have a whole lot of experience, but there are definitely companies that I saw and had a chance to put money in which will ultimately become successful or become a unicorn. Conversely, there will be big hits, like my first fund (I didn’t lead the deal) did the Series A of Datadog, and that became a gigantic win. As well there are deals we were shown that, maybe, we should have said “No” to and we didn’t see the red flags or we saw them and ignore them. Hopefully, we learn from them but certainly, we wrote those deals off as they weren’t successful. I don’t think it’s a long list of folks in my anti-portfolio, but there are definitely ones where we had the chance to win big.
The way that my fund makes decisions has not changed. Our approach is to try to be intellectually honest with whoever is bringing the deal to the table, to shoot for a unanimous decision. I consider it a function of a good corporate culture. We’ve known each other for a long time, worked together, we trust each other. Not being face-to-face means that body language which is a part of the equation is minimized; that’s potentially a challenge that we overcome by good corporate culture. The idea of not breaking bread with the founder prior to doing it changed. I’ve met a founder back in June or July of 2020, and we’ve met in person almost a year later. So, it is structurally different but in terms of the decision process and the diligence process it’s no different at all.
COVID resulted in a shift of resources: certain things that were priorities before and getting spent on before are not while other things, like secure communication (because now you’ve got people working from home), end-point protection, mobile device protection became more important. Some companies did very very well because of that shifting in the allocation of resources. On the other hand, other companies that were selling 6-figure, 7-figure deals but weren’t that a priority now is on fire, they are busy trying to do something else. While enterprises didn’t spend as much, some of our companies that have governmental contracts did amazing, because the government wrote gigantic checks for a while, buying these companies plenty of runways to ultimately get to enterprise and be successful. COVID certainly impacted us, but it impacted us both ways – something’s got slower and something’s got faster. Could we have foreseen this shift in spending? I don’t know. But when the dust settles, we can play as effectively as we can in terms of guarding our companies.
I am a big book reader which is, maybe, unusual and doesn’t exactly provide you with up-to-date information: what’s in books is at least several months old, potentially. I read about AI, behavioral psychology. I do keep track of who funded what deals and I do have the number of sources for that. There are several interesting sources for AI information, there’s a particularly interesting one on Chinese AI – it doesn’t help my investing per se but I love оust keeping track of what’s going on in the broader world. In terms of cybersecurity, we’ve got an amazing network of advisors, CTOs from the Fortune 100 companies that we talk to frequently. And frankly, I much prefer to get information from those folks than to read about it in a some kind of daily summary of cyber news.
AI Superpowers by Kai-Fu Lee, which is now a couple years old, is interesting. The Atlas of AI: Power, Politics, and the Planetary Costs of Artificial Intelligence by Kate Crawford is interesting. A Thousand Brains: A New Theory of Intelligence by Jeff Hawkins is interesting. I go back as far as The Fourth Age: Smart Robots, Conscious Computers, and the Future of Humanity by Byron Reese is an interesting book. If you want to think about it from enterprise or corporate, Amy Webb wrote a book called The Big Nine: How the Tech Titans and Their Thinking Machines Could Warp Humanity and it is about who controls the future of AI and how they are trying to buy it. Just a handful of ideas.
The companies that are going to change the world out of machine learning haven’t done that yet. I think we’ll be short-sighted. Talking about when the mainframe showed up – IBM was capturing the mainframe market. When network computing showed up – Cisco and its routers changed the world. When the internet showed up the initial browser companies, like Netscape, were harbingers. they were signals that the things that we anticipated were coming. Right now we are at the scaling of the horizon for those harbinger companies.
There are a number of companies that raise a whole bunch of money and then ultimately failed, but I think the reason was that they weren’t very promising – they just appeared to be very promising. Companies that I am thinking of in particular companies that failed the most staggeringly are companies that failed to fully understand their own unit economics, they were essentially wrapping dollar bills around their product and sent it to the doors of customers, and that just didn’t make any sense. Everybody said, “Oh my god, you’re growing like a weed that’s fantastic use more money,” and didn’t realise that in order to be growing like a weed it was costing them more plus more. When ultimately you try to normalise the spendings and get to current normal earnings power, regulate your income statement, it’s just impossible anymore. That’s when companies like The Country’s Best Yogurt massively over expanded, because I didn’t understand the unit economics. The retail company Fab in a number of years back burned out hundreds of millions they were spending too much. Those companies were like that from the very beginning, people just never noticed this.
Patience is one. Intellectual honesty is another – being able to look at facts with fresh eyes all and every time. Not have a preconceived idea about something what it is, what I can do. Being open-minded. Being able to defend an idea, but also understanding that there are multiple ways that risk can be measured and you might not be looking at the best way. Hopefully there are some of the same qualities in venture capitalist that there are in founders – great determination, humility, curiosity, sense of humour. All those things are really important for what you do everyday.
The goodness about venture capital is it uses both sides of your brain – the analytical side and the process driven side, but it also uses the creativity side, uses joie de vivre and wonder about what that could possibly be. I suppose there are games out there and folks who can see several moves into the future and understand what the back-and-forth will look like and how things might develop. Maybe that looks like chess, I don’t know, maybe it looks like bridge, maybe it looks like Texas Holdem – that’s a risk-return game and that’s a decision-making game with limited information. But I do think that there is a lot of different aspects and the better VCs employ many of those angles and aspects to do their job as vell as they can.
The goal of your pitch is to get one more meeting, that’s number one. The other thing is: technical founders are so smart that something inside drives them to show how smart they are, but all you need to do is communicate the information. The person on the other side of the table will understand that you’re capable of doing it – just make sure that the things that need to be communicated get communicated.