Sebastian Duesterhoeft of Lightspeed Venture Partners joins Nate to discuss Are Software Businesses Defensible? What is the S curve and How to Time it, and Lessons from Investing in Snowflake, Databricks, and Confluent. In this episode we cover:
- Investing in a Challenging Market with a Growth Investor
- AI Technology, Its Impact & Potential to Disrupt Industries and Create New Opportunities
- Investment Strategies and Price Dynamics in the Startup Space
- Investing in Enterprise Software Companies, Including S-Curve Analysis
- Market Size and Potential for a New Endpoint Security Player
- Determining the Size of the Queue for Cloud Security and Software Development
- Market Sizing, Vertical SAAS, and Software Margins
- Investing, Brand, and Distribution in the Tech Industry
Guest Links:
The hosts of The Full Ratchet are Nick Moran and Nate Pierotti of New Stack Ventures, a venture capital firm committed to investing in founders outside of the Bay Area.
Want to keep up to date with The Full Ratchet? Follow us on social.
You can learn more about New Stack Ventures by visiting our LinkedIn and Twitter.
Are you a founder looking for your next investor? Visit our free tool VC-Rank and we’ll send a list of potential investors right to your inbox!
Our guest today is Sebastian Duesterhoft, one of Lightspeed’s newest partners on the growth team.
General Partner and growth investor at Lightspeed Venture Partners. Sebastian recently joined Lightspeed from Coatue, where he was a General Partner and led investments in some of the of the most iconic companies such as Snowflake, Snyk, Gitlab, and Confluent. Prior to joining Coatue, Sebastian was also an investor at Silver Lake and Morgan Stanley.
Sebastian, welcome to the show
0:27
Thank you for having me, Nate. Of course, I’ve really been looking forward to doing this one. I’ve read a lot. Maybe let’s start with your background in your winding path to Vc. Can you share a bit about that? Absolutely. Yeah, it’s been a it’s been a winding path. So I grew up in Europe and Germany originally ended up in in investment banking in London for a couple of years had nothing to do with technology. Spend time there on on an m&a team that did everything from healthcare to airlines to mines, oil and gas, and everything in between. And then Kennedy just lucked out and fought my way to win that sandbag to Menlo Park. And that was my first touch point with technology. I actually had wanted to go to New York as every decent investment banker would want to, but couldn’t find the right team. They sent me to Menlo Park. It wasn’t what I expected. No tall buildings, it turns out over there. But it was really lucky. It was a really lucky turn of events. That team is fantastic over there. That was my first touch point with the technology. And then I just lucked out a couple of times fought my way to civil Lake spent about five years in really the late stage investing world from there with the CO to spend another five years there. That was really my first proper touch point with growth investing, sort of in the venture sense. And then then, you know, earlier this year, I joined joined Lightspeed ventures. So sort of gradually make my way earlier. And you know, to the venture venture style investing side, I guess, well, congrats on joining Lightspeed very exciting. So you’ve been investing for about a decade now over a decade ago. So I’m curious to mind those 10 years of insights, perhaps starting in reverse, a lot of people are saying that now is the best time to be investing. But before the show, you mentioned that this is probably the most challenging environment for new entrants that you’ve personally experienced. So I’m curious to get your take on that sentiment is now really the best time to be investing. Yeah, I think it’s it’s a, it’s a super loaded question. And one I think a lot about I don’t think I have, I certainly don’t have the answers to it. And there’s a few different ways that we could take it right. If I think about it more from a investment opportunity perspective, I think, if you’d asked me this question, six, nine months ago, I probably would have been a lot less optimistic than I am now. I think you put yourself into, you know, early 2022. And, you know, obviously, we all know what was going on in the public markets, you know, what we’ve just gone through on the private side, just purely from a valuation and share price standpoint. But even when you just sit back from that for a second, it’s also hard to deny that we were looking at a world where a number of the the mega trends that we all had been investing behind was definitely they were all, many of them were, we’re nearing sort of the maturity stage or later maturity stages in their s curves, right. When we think about, you know, an enterprise investing, certainly, that transition to cloud. You know, when you think about mobile, and in consumer, a lot of those things that we’ve been kind of mining for investment opportunity for the prior decade or so, they were just coming towards the later stages of the S curves, and it wasn’t necessarily apparent what the next thing was, was going to be. So I think, you know, viewer, again, sitting in early 2022, it felt more challenging. I think we’re all very thankful that, you know, we now have aI again, and it’s not the end all be all to everything. And there’s there’s lots to be skeptical, but there too, but I do fundamentally believe that it’s opened up a ton of really interesting opportunities, long term, really, really excited about it near term, we’ll make some of the same mistakes we will always tend to make in this industry, we always get a little bit overexcited about these things. And that’s okay, that’s part of our industry. Right. But I think if it wasn’t for that, I think lots of us would have a harder time really pinpointing specific areas or trends that we’re particularly excited about. Right AI is driving a lot of it, it sounds very stereotypical, but I think it’s true. So that’s the investment opportunity side. And then I think that the other sort of the other side of it, which is more what’s happening, you know, with investing firms, it’s a whole different set of questions and one that’s that’s really complicated. Again, I have some thoughts on it. It’s certainly not fully formed. And, you know, we all know that that as an industry, we’re going through a phase where we’re digesting a lot of things we all did and 2020 2021 Not all of them were things that we would do you know, the same way again, today we’ve all grown the firm’s in very meaningful ways, both in team size and kind of capital volume that
5:00
We now have available. I think there’s lots of questions around, you know, what the right font sizes for for some of these asset classes were in, you know, are we going to see, you know, 510 $20 billion growth funds? Again? I? I don’t think so. And that has a lot of sort of downstream effects on, you know, what these firms I may have to look like over the coming years right now. And I think we’re just in the middle of digesting that. So lots of things going on there. Yeah. I,
5:28
I wanted to maybe pick your brain a little bit more on the technology piece of what you mentioned, specifically AI? How do you compare the AI cycle to past cycles, such as Cloud, as you mentioned?
5:41
Yeah, it’s interesting. So, you know, before we came on to the show, we talked a little bit about differentiation and modes. And, you know, we talked a little bit through, we’ve all read seven powers, we’ve talked through some of the, the elements of that, and we touched on counter positioning, which is one of the seven powers. And that was one of the critical elements, I think, that allowed new entrants to really take over when when subscription models and cloud took over, because it was just, you know, it was a difficult switch for the incumbents to make give up the perpetual license model, etc. So that really allowed a kind of positioning more for new entrants to come in with that subscription model. And if you now look at AI, I don’t think that exact same kind of positioning benefit exists, which I think is one of the reasons why, you know, I I think that that, and again, that’s I think, the thinking that that many people seem to be sharing right now. But I think that the incumbents in this kind of technological change are better positioned than they were, you know, when the cloud came about. So, you know, that’s something that I think about a lot, it doesn’t mean, there’s, there’s not going to be really interesting opportunities for new entrants to emerge and your great companies to be built. But it’s going to have to be, you know, different. It’s not reinventing an existing workflow, you know, with AI embedded and in AI enhanced, that’s probably not going to be enough, I think we’ve already seen that the incumbents have moved faster than they probably ever have almost before. And the technology has also been widely accessible. So it’s, it’s also not one of a of a corner resource, where we’re only a handful of folks have access to it, but it’s very widely addressed, widely available. So I think as a result, the areas that I found more interesting, as you know, in, you know, where I can be leveraged to maybe go into verticals that otherwise weren’t really addressable, where they can really reinvent the way that some of the processes and workflows function in those verticals. And it’s just a completely different kind of behavior and model than existed before. You know, I think there’s definitely going to be opportunity around sort of a new emergent infrastructure stack, I think every time you have a new data model, a new way that data is being leveraged, it happens to be, I think, a new infrastructure stack that’s being formed around that. So I think there’s going to be elements of that there’ll be interesting, but yeah, I do think it’s quite different. I do think, long, long term, this could be even bigger than cloud. And I think it should be bigger than cloud cloud in a way was was shifting over existing spend into, you know, an increase in spin, I guess we’ve all seen that the cloud versions of those companies have gotten bigger, and then the on prem versions of that same software, that at least, I think anecdotally that tends to be true. But the way that I think about it is that all software at the end of the day is about shifting, spend for a given workflow from labor spend to software spend, right? By automating parts of that workflow. And, and, you know, it always, and I think AI has an ability to capture a lot more of that of any given workflow and shift that over to a technology spend versus labor spend, then pure software could in some ways, the level of automation you can achieve with AI, I think is substantially in particular, obviously, these these newer forms of AI that we see now. So as a result, I think that the you could make an argument that the value that can be created through through general AI over the long term should be or could be bigger than what was created with with a cloud transition.
9:31
One of the things that have been hearing anecdotally at least that has been interesting, or some of these esoteric spaces where cloud hasn’t had the same demonstrable outcomes, as you know, five 10 billion call it you know, no, no large publicly traded companies, such as LegalTech. For example, there there are solutions now that are being adopted quite rapidly in the spaces that have been traditionally very difficult to penetrate. by your traditional software companies, and now with velocity are being penetrated by software companies utilizing generative AI. I’m curious, like as you take a look at spaces that maybe were not as ripe for disruption, or maybe not as attracted from an investor standpoint, because they can’t be penetrated with velocity and grow 3x year over year when they’re at 50 million Arr, to jump to call it 100 million and beyond, does has this shift? Or is this cycle, changing the way that you’re thinking about certain markets or certain industries? And what might be attractive today? Or could be attractive a few years from now, that wasn’t attractive five years ago?
10:45
Yeah, I mean, I think legal is a perfect example of it. It’s one that we’ve spent a lot of time I’ve personally spent a lot of time in. And if you’d asked me that same question two or three years ago, you would have probably gotten the answer. And visa, it’s not that interesting, right, there’s just not enough money to be made, you’re going after a fairly fragmented set of customers, the value capture for each customer is fairly low, even if you’re the most integral software solution, you know, it’s it’s, it was still oftentimes the ACV, where low single digit 1000s of dollars per year, that makes for a pretty difficult go to market model. And as a result, one that is hard to scale, you don’t tend to see these hyperscale sort of growth opportunities that we all like to see in venture. And we have seen that, in particular, legal, general AI seems to be changing some of that. And it goes back to the point I made around generative AI being able to capture more of the value of a given workflow than pure software code. So there are examples of companies that, you know, if you, for example, looked at the practice management software solution that that same law firm would use, they might be spending one or two, three or $4,000 per year on that practice management solution, which that would have been perhaps the core software solution for that organization, you know, the past would be their CRM into the scheduling engine, it is very critical to their, your day to day life, but it just didn’t capture a lot of value. And now that exact set, the exact same customer might be spending, you know, 3040 $50,000, on some of these generative solutions that are out there that might help them to summarize vast amounts of data, search across vast amounts of data, generate first drafts of documentation that’s required in the legal space. So the value capture is a lot higher. And I think it’s sort of a perfect example of this technology, being able to shift more of that labor span. And in legal, obviously, we all know, it’s pretty expensive labor spend moving that over to technology for any given workflow. So I think it’s a perfect example of that. I also think that it’s sort of like a good example of how, you know, Gen ai plus human in the loop, I think is the right solution in many industries today, because, and you will probably get to it as well. But I do think one of the challenges right now and why I think in summary, is we’ll probably find ourselves getting a little bit ahead of ourselves and Jenny, is that the products aren’t 100% there yet, right? Like, I think we all know that they’re, they’re incredibly impressive, like, the 08 is incredible. Like, it all blows our minds, right. But when you then shift from, this is incredible, too, can I use this on a day to day basis, and like, let just let this thing just run my life. It’s not quite there in many cases. So if you have these use cases where you can actually still capture a ton of value, you know, by augmenting an existing human with this technology, and you still control the overall quality of the output with a human versus trying to have the machine do everything end to end, then I think it becomes really interesting and illegal. I think that that that tends to be a use case where that works really well. One other spaces,
13:59
aside from legal have you guys been looking at? Or are you leaning into? Or maybe you’re not actively looking at but when you’re thinking creatively about where things could go over the next couple of years? What markets can be interesting over the next couple of years? What comes to mind as in terms of what markets lend themselves toward venture opportunities that didn’t in the past?
14:23
Yeah, and, you know, some of these I personally haven’t spend time in so I can speak as deeply about him. But I think sort of one simple framework is certainly around, you know, when you when you think about what made legal or what I think makes legal suitable, it’s the huge amount of unstructured data that exists in that space, obviously, typically just text documentation and the need to summarize across it or find something in it, the huge amount of tax that needs to be generated, but then also the, the, the individual that’s so expensive in that area. And if you think about some other areas So that applies like, certainly healthcare is one. I do think that’s super interesting. I haven’t spent enough time myself to point to very specific examples. But I would say like the the framework, it’s not even a framework, but a little bit, the current characteristics I described, that apply to the legal space probably, in many ways apply to the healthcare space as well. So I’m sure there’s going to be interesting things there. And then over time, it will go everywhere. I’d love to find something at some point in, in insurance and construction. I think there’ll be the there’ll be interesting things, but probably my hunch would be healthcare and legal are probably one of the one of the early ones where you start to see interesting companies.
15:38
One other thing I wanted to get your your insight on, as you reflect over the last 10 years is your relationship with price? Like how is your relationship with price changed over the past 10 years, as your career has evolved?
15:53
Yeah, it’s a tough one. And I’m glad that you that you you call it price and not variation. I think it’s, you know, like in investing, there’s obviously I think there’s always an element of this where you have to play the field that’s in front of you to some extent. And that means that you’re going to end up paying a price versus paying the valuation that you think that company should have at that moment in time, perhaps. And I definitely had to learn over the years, like it came from a place that was private equity, or came from a place that was more public market oriented, which I think really helped me because you, you develop a good understanding of how something should eventually be valued, you know, at some more and more steady state when growth slows down when the margin structures should should stabilize in some fashion. So I think it sort of grounds you in, in, in some of the fundamentals of how valuation should work over a long period of time. And I think that’s very important. And it’s has helped me a lot. But it also is something that you have to be able to not remove yourself entirely from, but you have to get comfortable with those models, not necessarily applying the earlier stage you go. It’s not not to say you should forget about them. But it’s just you have to open yourself up a little bit to be able to lean in to lean into situations where you’re clearly paying very high relative valuations, relative to the to the revenue at that moment in time. But you have a perspective on what that company could become, you know, what the opportunity size for that company is, that allows you to lean in in ways early on that that do feel uncomfortable, but that are justified relative to the outcome that could be achieved some number of years down the road. And there’s no question that in, you know, 2020, and 2021, we massively including myself massively got ahead of ourselves and told ourselves, we could pay these leanin prices for many, many, many companies, when the reality is there’s a very small set of companies that you should be able to do it any given year. And I think as as an investor, like I think about it as what I do that I need to have incredible amount of conviction, I need to understand the space incredibly well, I really have to have a view on what that company can become over a 510 year horizon, and I need to have extremely high conviction in it. And I need to be aware that, you know, paying these 20 3050 or 100 times ARR multiples, means there’s a lot of things that have to go right. And there is something to it, obviously, that the later stage of the company is the harder it gets to pay these kinds of valuations like now looking back at some of the things that we did, when companies were a billion plus 2 billion $3 billion of valuation and the multiples 100 times that current AR, that is such a difficult thing to make work. It’s incredibly hard. And there’s probably I cannot even tell you an example, actually where that’s worked out historically. I’m sure there is one, somebody will correct me on the rest one, but it’s extremely rare. It’s a little bit easier, in some ways. If you look, you know, look at the sub $500 million absolute valuation, because again, then then sort of the ratio of valuation relative to the opportunity size relative to tam tends to be a little bit easier. But we’ve certainly got massively ahead of ourselves in 2021. In particular, there’s instances where we paid these extremely high multiples for fairly late stage companies.
19:25
Do you think things have come all the way back down to reality yet, or our price is still a bit inflated?
19:33
I think we’re in a weird spot. I think in AI, we’ve we’ve all seen behavior that in some ways, reminds us reminds us of 2020 2021 I would say there’s also this this interesting dynamic where you have a lot of strategics playing very meaningful roles in that area that in ways that wasn’t really true in 2020 and 2021. Right you have the Googles and Microsoft and Nvidia As of the world investing into these companies at, you know, valuations that are very hard, I mean, even even from an from a, from an aggressive investor standpoint, those valuations seem very aggressive. It obviously has something to do with the fact that it’s not necessarily in a financially minded, more financially incentivized position, they’re taking in those instances, right. And like that is driving obviously, some of those valuations higher as well and has an impact on the environment. But then it’s also, at least at the what I would have historically called the core growth stage, you know, call it a $25 million dollar AR company and up, there just hasn’t been a whole lot of volume, in terms of number of companies that are trying to raise this year so far, I would argue, and part of it, it has to do with the fact that lots of them raise in 2020 21, huge amounts of money at high valuations, they haven’t been able to grow yet into that valuation of full extent. So if they were to come out now be a difficult conversation that we had with you can raise that five valuations or down or what that might look like, they also are, a lot of them still have a lot of capital. So they don’t have the extreme urgency to raise, many of them have been able to extend the runways meaningfully. So they don’t need to raise. So the few companies that do come to the market, I think they’re they just then facing a situation where you do have a dozen or so growth investors that by and large, have been relatively inactive for the last 18 months. Now there’s finally a growth asset. And it’s sort of a supply demand imbalance that still helps these valuations be fairly healthy, I would say, I think that might change over the next two or three or four quarters, many of the companies many of the 400 software, unicorns are three to 400, whatever the right number is that raised in 2020. And 2021, they’ll they will have to come back eventually. And the moment you start to see 1015 20 companies in that sort of core growth bucket arrays in any given month, the supply demand will be a little bit more balanced. And I think then we’ll start to see, probably, you know, in some ways, more reasonable multiples, you know, we’ll probably start to see a lot more done rounds. You know, it’s yeah, that’s
22:09
Yeah, it feels like we’re going to see more situation similar to hop in caught it as well. Right, that feels like it’s bound to happen.
22:17
Yeah, I think it’s, it has to right. Look, it’s again, I don’t know what the exact number is today. But it’s probably somewhere in the order of 4400. Software unicorns, if you look back at, I think the average number of tech IPOs tends to be 40. Every every year, and like obviously, not the same, you have some years at a higher, some that are very low, but over the long term sort of averages out to 40 I think in the US generally. And that’s not just software that’s across software and consumer and fintech, etc. So I guess, you know, we’d have 10 years worth of backdrop right now, clearly, it’s not going to work out, right, like not all of these companies are going to make it the public market, probably some large share, it’s not going to make to the public market, somehow this whole thing has to unwind a little bit. And it’s going to happen, right? And it’s going to be difficult in a lot of in a lot of places for sure. Yeah. And I think we’ll start to see that happen. Probably, you know, throughout next year, it’s my guess.
23:10
Yeah. Well, I wanted to talk about how you invest in these companies, because you’re in some very notable names I mentioned at the very top of the show, but snowflake GitLab snick. I, you have an eye for picking up. So I’d love to walk through your framework, specifically how you think about investing in these enterprise software companies. And maybe first, we can start with timing, and then we’ll get into some of the other buckets as well. But you mentioned this concept of S curve, and how critical that is in timing the market and when the market is receptive to a startup solution. So I guess first, can you describe the S curve and how this factors into your decision making as an investor? And also how do you measure where a market is at? relative to its s curve?
24:04
Yeah. Yeah, at the highest level, it’s, you know, S curve, what we mean via this, it’s, it’s sort of a widely used model for for market adoption. I think we’ve probably all seen these these graphs right? You have this, you know, pretty lengthy flattish part of the curve, where you have some early adopters adopting a solution, you know, not, not a lot of total revenue is being generated. And then at some point, you enter this extremely steep part of the curve where a really large part of the market adopts a solution and then eventually you get into the latter part of that curve when it flattens out again, and you It sort of takes a very long time to get the laggards to adopt, you know, a solution as well and make most most technologies tend to follow that kind of curve and in some fashion, right, and sometimes you have moments when that really steep part, you know, gets incredibly compressed. So right, like zoom to COVID would be a prime example of that, you know, sometimes it takes a little bit longer, right. But generally, that’s sort of the curve that we that we all think about. And as a growth investor, like, for obvious reasons, the perfect time to invest will be basically just before this turn turns, this curve turns really, really steep. So we try to do our best to identify when that’s the case. And, yeah, we don’t always get it right. Certainly, but but we try to, and I think, you know, from a growth investor perspective, you know, some of the things that I look for in when I look at a particular enterprise space, it’s, it’s pretty obvious things, but it’s things like, you know, you talk to a company, you hear about the the levels of ACV, they’re able to get out of companies, right? Like, it’s just sort of starts to signal that companies are attributing real value to the product that’s being offered to them, right? Is it something that can start to extract 50 100 150 $500,000? A CVS, right, you start to look at the types of customers that are starting to adopt it? You know, is it just sort of the kinds of companies that you would associate with, you know, early adopters and leading edge type of companies, right? Is it mostly exclusively Silicon Valley startups and sort of, you know, friends and family type companies? Or is it actually starting to reach the broader broader economy, folks that are far away from Silicon Valley? So you started looking at that. And I also think that oftentimes, a given trend, or a given company that you’re going after is tied to some underlying trend in some way, where you really want to pay attention to that underlying trend as well, I think we’ve you and I have talked about cloud security before. That’s one that’s super tidy, super closely tied, obviously, to the adoption of cloud cloud infrastructure and an effect. So if you observed the curve of adoption of cloud infrastructure, you could have probably seen that over time, that would necessitate security solutions to protect that, right. And this is not a perfect analogy. But typically, you’d say, medical, some folks would say that security solutions tend to follow infrastructure with some amount of lag, maybe it’s two or three years or so you’d kind of just have to see a a certain infrastructure component be widely enough adopted, that some of the adversaries start to focus on it and start to care about it. And you know, of course, that’s when when companies and also start to think about well, now we need to protect it, how do I do it? And you know, that’s what what would happen there. I think similar with AI, you will see something very similar. Now, right now, it’s all about just finding ways to adopt these models in, we started to see companies pop up around AI securities, well, it’s probably relatively early from a growth investor perspective. But I’m pretty convinced that in two or three years down the road, we’ll probably be talking a lot about AI security companies that are able to protect against prompt injection and things like that, right. But sort of a new new security threat vector that that is created because of the underlying trend, that sort of moving up the S curve.
28:11
Yeah. You mentioned how we were talking offline about like red lock and evident IO for which are both cloud security companies just for the audience’s context in how they were they were both I believe they were both acquired by Palo Alto Networks. But they missed the real opportunity, obviously, that was in Oregon and others capitalized on are there instances that fly in the face of red lock and evident IO, which those two pave the market educated the market, or they were a bit early, they didn’t have the first mover advantage is ultimately what I’m getting at here. But conversely, are there instances in which you’ve seen a company have the first mover advantage in which it’s advantageous to be first? And I guess, how do you think through situations in which being first is actually better? Versus you’re gonna wait to let the market develop? And it’s actually not better off to be first to market?
29:10
That’s a great question. I don’t know, I can’t think of a great example, where you had to be the first mover in a particular market. I think we can probably think of plenty of examples where the second or third player had still had a very, very good chance. Yeah, I don’t I can’t think of a good example actually. And I don’t know that there isn’t a session necessarily. I think it goes back a little bit to the conversation we had around you know, moats in software and auto software like this notion of like cornered resource that that you know, in some other industries might apply. And if you were the first one to to corner that resource, nobody else can follow you. It doesn’t tend to exist in software all that much. I think the foundation Moreover, companies have been trying to make that argument. It felt like every week, you’d be talking to a company that had you know, the five only people in the world that Do this. And the next week was another five. Now we’re sort of, you know, this generation of a coordinate resource for some for some period of time. I don’t think that exists typically all that much in software. So I, I struggle a little bit have come up with a great example of you had to be first in the market to win win a space,
30:21
I can’t say that I can think of one off the top of my head, either. I think there are certain models and maybe lend themselves toward moving first or moving faster or earlier helps, but they usually touch something that’s discrete.
30:38
Yeah, I think it’s also within within reach, like, you know, if you look, for example, at a crowdsite versus a sentinel one, I don’t know what the exact sort of timing difference was between the two of them emerging, I actually want to say send an A one maybe even either been earlier or within a year or two of CrowdStrike. I think most folks would agree that at least in core endpoint, it’s proven to be it’s proven hard for Sentinel one, to really catch up with CrowdStrike.
31:10
Yeah. Do you think that is the way I think
31:13
it’s probably a combination of a product that is probably insufficiently different from CrowdStrike. So there wasn’t really a big enough reason for for, for sending a one to emerge as as sort of like next generation or something that would disrupt CrowdStrike, there’s certainly an element around CrowdStrike, owning many of the channel partnerships, and maybe not owning, but certainly, having very well positioned with a lot of the channel partnerships in the industry and insecurity, channel partners tend to be very, very important. And then I think it also has to do with the underlying market and endpoint security, I actually think is sort of one of these good examples of we’re probably further to the right on the S curve than we are to the left. And the further right, you you come on the S curve, the harder it’s going to be for a new entrant to come in, unless you really have a very good reason to dislodge the existing player in that market. Right. Again, I think there’s there’s far fewer at bats in endpoint today, meaning there’s far fewer companies that still need to switch over from Legacy endpoint, a next gen endpoint, there’s far fewer companies that have nothing in place on the endpoint side than there were five years ago, right? Like CrowdStrike certainly benefited massively from being able to transition so many people over from Symantec and McAfee sort of that, you know, 1.0 endpoint to next gen one endpoint that really helped them go up that S curve really, really quickly, where sending a one came much later. And it’s now sort of competing at a moment in time when that market is much later on that S curve. And the incremental dollar volume of new deals that come on, you know, come to the market every years is getting smaller and smaller, I think.
33:03
Yeah, that makes sense. So I guess, let’s assume that we nailed the timing of the market for a potential investment. And we’re convicted in that next bucket. I’m curious to get your take on and in terms of how you think about is market sizing. If it’s broadly speaking, price times quantity, at the end of the day gives you your market size. And I view and I’ve talked about market size a bit offline, but I know quantity is more challenging to estimate than price. I’m curious to hear how you think through quantity? And what quantity could become why investors miss. On what Miss on quantity, specifically, in that p times q equation?
33:53
Yeah, I think queue is difficult for for a bunch of reasons. There’s, I generally, you know, when we focus on enterprise software, in particular, I think the the cues that are a little bit easier to determine are seat based models, right? That’s something where if I thought about something like a good, good lab, for example, the number of developers out there is fairly well defined. You know, it’s probably somewhere in the order of like 11 or 12 million developers and professional organizations, there’s lots more that, you know, sit outside of them, but we think about, you know, organization that will likely be paying customers, it’s maybe something like 11 or 12 million. It’s like a fairly defined market, then there’s lots of other questions around, well, how much of that queue is actually truly accessible to you? How much of it doesn’t already exist with another nextgen vendor in this place? This case like GitHub or so how quickly will some of the remaining seats turn over? Like there’s a lot more kind of down some questions you have to ask once you’ve determined the queue, but then there’s also other companies where The Queue itself can actually be somewhat, like much harder to determine, like, what is like the what is the queue for snowflake? It’s much more complex, right? And it can be done, I think folks have come up with with interesting ways to quantify that, Tam, but it’s, it’s not a person, right? It’s quite hard to think about, you know, how many queries a given person might write? Right? I don’t have a good concept for that, like, well, I have to, you know, write 100 SQL queries a day. And, you know, how many of me would would have people like me would have that volume of queries, like, that’s not a framework that’s like, easy to access? Right. So So determining that crew queue in that case is a lot lot harder. And there’s, there’s lots of lots of examples like that. And then what I found is that or similar similarly with with cloud security, right, like what is what is that queue, maybe it’s it’s workloads, right? Like, but it’s, it’s simulated, sort of like a harder, harder cue to get out, then sort of the seat base, for example. So what I’ve typically found to be quite helpful is, you can oftentimes find fairly simple kind of frameworks for a given market that are actually quite powerful. And they also can be really helpful in making sure that you, you stay open minded on the upside that might exist in a space, right? Like, it’d be very easy to look at evident IO and read log back in the day and say, well, like, I mean, they sold for $500 million, right, or 400, or 300. Like, clearly there have not been any big outcomes being created in, in, in cloud security, right? Like, clearly, it’s not a big market. And maybe at that moment in time, you know, large organizations, were only spending $100,000 hcvs. And would have been easy to say like luck. ECGs are pretty small here. It’s can’t be that big of a market, if that’s true, right. But then if he had stepped back and said, Well, historically, most security categories can generate somewhere on the order of five to 10% of the spent on that underlying infrastructure. Right. So if you looked at the networking space, you’d have a certain dollar amount dollar amount that’s being spent on the actual network infrastructure. And you would have said, well, people are willing to spend maybe five to 10% of that, on actually securing that infrastructure. And if you believe that, that would broadly hold true for the cloud infrastructure world, you would have said, well, you know, the spend that’s happening on AWS and GCP, and Azure today, or even three, four or five years ago, it’s pretty, pretty darn big. And as we all know, is so growing really rapidly. So if that equation even remotely holds true, and even if it’s, instead of five to tenants, two to five, surely, you would see a much, much bigger company. Right. And that’s how folks started to think about it. And obviously, now fast forward, and, you know, folks like the wizards and archives and lace rooks and Palo Alto Networks in the world, you know, the ACPs, for the cloud security solutions are well into the millions, if not into the double digit millions. So certainly the ACB question has been answered. And we’ve obviously also just seen much, much bigger companies now. Right. And it’s clearly I think, viewed as the category that is probably the fastest growing security category at the moment. Right. So sometimes these pretty simplistic frameworks can really help you make sure look, you know, let me make sure I’m not missing something here. Yeah.
38:30
What have been your biggest misses when it comes to market sizing? Is it miss calculating the queue part of the equation? Or is it something else?
38:40
Yeah, it’s the queue, usually, I would say, right, I think it’s, it’s, it’s oftentimes in I find that it can oftentimes be in areas that we just tend to spend less time in, as venture investors, like, they oftentimes can also be further away from sort of the traditional venture ecosystem and West Coast ecosystem. So you just don’t you just not quite as close to it, right. Like, I certainly got penggiling. Very, very wrong early on. And that’s a whole different conversation, because it’s vertical SAS and I think lots of folks have made mistakes around vertical SAS, including myself. But it was also an example of I probably didn’t understand and know that space well enough to really have a perspective on on where you could go with a product, what adjacencies you could cover, and ultimately how much value a software solution could capture. Right? And yeah, so I think a lot of it is a lot of it or getting it wrong is usually Q and and oftentimes it’s tied to not understanding a current category well enough. And then oftentimes, you know, another example would be, you know, SMB, SAS, it’s obviously been it’s been talked about a lot. SMB was historically something that lots of folks are very skeptical off in terms of can you build big companies there and it was both a combination of tam but then also more so a little bit. Whether you’d be able to address and actually capture that theoretical time and in sort of economic in an economic fashion. And I think that’s an area that generally lots of folks, including myself have gotten wrong, right. And I think part of it is under estimating how you could go from owning one piece of an SMB two, turning yourself into a pretty deeply ingrained operating system for that particular SMB vendor, particularly if it’s in a vertical, how you might jump from that starting point to other players in that value chain around that vertical space, or that SMB space. And those are things that lots of folks have underestimated.
40:40
Yeah, well, they’re rare, too, though, right? I mean, there’s a handful of vertical SAS companies that are valued above five or 10,000,000,005. Toes, viva, I mean, handful, so
40:52
it’s still not a lot, right. And I still, I still genuinely subscribe to the notion that that horizontal markets are bigger, right? Again, you look at Salesforce Salesforce itself, the horizontal kind of CRM is, you know, meaningfully bigger than any of the verticals that exist, like Viva for for healthcare, etc. But again, it also doesn’t mean that you can build, you know, 510 15 20 billion public market cap companies and verticals. And that’s certainly sufficient for us to achieve venturous outcomes, right, obviously, all depending on when you when you when you invest in these companies, but there certainly are good examples of great venture investments to be made.
41:30
Well do. I mean, the next topic that I wanted to talk about was terminal margin profile, and defensibility. And I guess a segue into this is, if you find that these horizontal software companies spend more on sales and marketing to retain their customers than the vertical software companies, because the vertical software companies lend themselves more toward a winner take all dynamic in that particular vertical, and they’re more ingrained in the workflow, as you mentioned. So they own more parts of the workflow, more stakeholders are involved in the customer, as opposed to the horizontal player. So I’m curious if like the the margin profile for the horizontal versus vertical SAS company, like, compare and contrast those two, I haven’t seen what that looks like at the growth stages. So be curious to hear if there’s discrepancies or not.
42:21
So, you know, ultimately, the margin structure I think for for the vast majority of software companies will predominantly depend on your sales and marketing spend. And we in software, with that one luxury, that gross margin tends to be fairly straightforward, at least in most cases, that’s that’s not that’s not the case in consumer and fintech. So that’s one that’s a little bit easier for us typically. So that’s good. GNA and r&d are typically fairly predictable, right, and you kind of know where we’re, that should end over time. So then it’s all down to, to sales and marketing. And that’s ultimately a function of the customer base you’re selling and how competitive that market is, right. And I think in, in some of the horizontal spaces, you you do face extremely strong platforms, the Microsoft, the Salesforce of the world, ServiceNow is of the world that make it extremely hard for for new entrants to overcome that distribution advantage that those planets have over time. And that’s, that’s why I think you, you see a lot of the more point product horizontal players struggle to get too strong margins over time, on the vertical side, or on the SMB side as well, I think you can get your extremely high margins, because in verticals in particular, to your point that competitive setup can be more favorable. In many cases, you also tend to have really sticky products, which is oftentimes a function of you being able to capture more and more value and and own more and more space within your existing customers which makes them more sticky. You also tend to upsell a lot into those existing customers and the upsell tends to come cheaper than the initial sale. So all of that helps you when it comes to driving sales efficiency and and ultimately margins in those categories. And then an SMB, you know, it’s that’s probably somewhere in the middle, you can have a similar advantage of going Multi Product owning that SMB in a really broad way, increasing retention as a result driving upsell, which comes cheaper than the than the initial sale, but then you also inherently deal with a customer base that’s, that’s Chaeronea then these large companies, right and that’s always gonna work against you in some ways. So you’ll probably end up somewhere in the middle between the two. But that’s that’s like very, this is a very broad topic. So happy to Yeah, yeah. But there’s a lot here.
45:01
Yeah, no, there is. So when you think about terminal marginal profile over time, how do you extrapolate that into the future? For what? Let’s pick an enterprise, a horizontal enterprise SAS company, I guess. And obviously, the competitive set matters and all these things, but I’m curious to hear you like the way you would think through the terminal marginal profile over time. Because as we talked about new alluded to this before the show, software companies and their modes are, I’m not gonna say non existent, but there’s research to be done might be the way. Yeah. So how do you think about the potential competitive set in some of these other variables that might not be at play yet, but they will be in the future, and that will affect the margin profile in the future? So how do you think about that? And how the terminal margin profile ends up revealing itself?
45:55
Yeah, yeah, we were talking about this a little bit before, right. It’s, it’s one of those areas that we’ve all read some of these books that talk about moats, right, seven powers and others. And we always, there’s always lots of examples of companies that, you know, you go back to the railroads, you go back to like a mine, you know, or something like that. And there’s all these great examples of how they have almost unsurmountable moats around, you know, some cornered resource or whatever it might be. And, and, and I always wish there was, there was actually, you know, somebody would write the book that apply, tries to apply all that to software, and I yet have to find it. And I think part of it is some of those traditional ways that your modes get created for businesses don’t necessarily apply quite as well to software. It doesn’t mean there aren’t modes, but I think they’re generally probably a little bit shallower. Overall, I think the the ones that do exist in my eyes are, there is some element of economies of scale, but maybe not in the complete traditional sense, but more around the distribution, the distribution economies. And I think that is very real, that’s definitely what you see, obviously, like Microsoft would be the perfect example of it. That’s why you know, folks like zoom and slack, as the horizontal entrance in those spaces, tend to struggle eventually. And then beyond that, I think the next next one, that’s, that’s certainly existence really critical, maybe it’s the most important one to some extent, as it relates to margins. And that was probably one that was even more pronounced in the on prem world, because it just required such a humongous effort to rip and replace something in the on prem world, it’s probably gotten a little bit easier in the cloud world. So you don’t actually have to manage that underlying infrastructure yourself. But I think I still think it exists, right? Like you are embedding the software solutions in your processes there, you’re teaching your entire workforce, how to use something that makes it inherently difficult to rip out. And, you know, that obviously, drives growth retention, which we all know is an important metric for businesses. And there’s, you know, pretty obvious reasons why the private equity players care so much about course retention, it’s because, you know, that allows them to really control margins, almost better than anything else, because push comes to shove, they can always reduce the spend on GNA, and r&d to nearly zero, and even reduce the spend on an s&m, really meaningful amount. But as long as they’re keeping the existing customers, they’ll be able to generate a revenue stream for a very long time. And that will become a really, really high margin revenue stream. Assuming you can dial back all the other spent right now you’ve gotten rid of your growth. But as long as it grows, retention is high, you have a very valuable screen. Right? And like, that’s, that’s what I do think lots of companies can actually achieve really high margin software, it’s just the question is, at what point are you willing to cut off your growth? And if you’re in areas where the stickiness isn’t, you know, in the high 90s, then you obviously have, you know, we have that working against you and you sitting on a melting ice cube?
48:58
My last question on the topic of defensibility was around brand, like, do you do you think about brand at all for enterprise?
49:04
I have to admit I, I don’t a whole lot. And I don’t know if that’s right, or not or wrong? I’m not.
49:11
I’m not surprised in that came up, like while we were talking, like, be like, specifically in security, almost like if you’re in the C suite, and you’re saying you’re secured by this brand. It’s a recognizable brand. If that’s a vote of confidence, I don’t know. But it’s something
49:30
a little bit of that. But it’s probably it’s probably in some ways tied to distribution actually as well. Like, you probably are more likely to have that brand. If you’re present in lots of places. And people think about you all the time and they’re going to think about you all the time of you know, the channel partners bring you up all the time. If you’re already using two or three or four other products like that may drive more of the brand. This probably there are probably more There’s probably brand more on things like dev tools, for example, like there’s definitely buyer types or user types that are probably more susceptible to brand made the developers probably one of them. The Silicon Valley knowledge worker is probably a little more brand aware when it comes to the UCaaS tool they want to use. I think the vast majority of Silicon Valley companies are probably still using Slack even though, you know, probably many of the traditional companies have decided they don’t need that brand. They okay using teams, but in Silicon Valley, I think people still want, you know, Slack. So this is probably a little bit of brand value, but I have to admit, I probably think a lot less about it than then I would in other industries.
50:43
Yeah, yeah. Interest saving Ron yourself. Especially if we could feature anyone on the show. Who should we interview and what topic would you like to hear them speak about?
50:52
Yeah, a few folks that I just look up to a lot and I respect a lot. It’s, you know, people like Dave Yuan, who was at TCV is now a tight mark. I think he’s one of the the absolutely most thoughtful people in investing. I, I am fortunate enough to talk to him relatively regularly. He’s awesome. You know, somebody who I think is still a little bit more under the radar, perhaps as Android Avenir. You know, he he’s, it’s so impressive what he is built with that firm over the last six, seven years, he started out building his own fund, probably mid to late 20s. And has built it in a pretty sizable for him pretty quickly. And it’s just sort of the the ultimate combination of hustle. But then also really, really deep first, and I think that’s, you know, he’s he’s another approach and that’s super interesting. Thoreau’s that. Arena Holdings is super interesting. I don’t think anybody’s ever gotten the amount of podcast, man. So that’s maybe that’s worth worth trying? Yeah, those are some of the folks like generally a, you know, I think in the investing world, I have, in particular, a lot of respect for folks that are able to move so fluidly across different asset classes as well, right, folks who can do early stage public growth stage, private equity. That’s, that’s really interesting to me. And then now here at lightspeed, I’m learning sort of a whole new lens, or I’m getting a whole new lens that people who come from the operating side have really helped build companies. And that’s really, really interesting to me as well, here at lightspeed, I think is fantastic, has an incredible ability to judge kind of founder quality and just something that that I’ve not seen before at places like civil code, it’s just not sort of what you do there what you learn there. It’s just a different different different DNA of investor and I think he’s he’s fantastic.
52:41
Yeah. And then last, what’s the best ways for listeners connect with you and or lightspeed? Yeah,
52:47
so I’m not super active on anything social media, I do have a Twitter account. I think I’ve maybe tweeted once, so it’s not super active. I’ll probably won’t respond there. The best way is probably email sebastian@lsp.com. And I’m happy to respond. But yeah, I’ve always debated whether I should be more active on on on some of these social platforms. But I I always feel like I’m not a very good fast twitch thinker. And that doesn’t lend lend itself so well to Twitter, I think, well, x, I should say, sorry.
53:22
Sounds good. Well, thanks again for doing this. And I look forward to having you on again in the future. This was a lot of fun. I feel like this could be multiple episodes.
53:32
Like Awesome. Thank you for having me. This was fun. Of course.
Transcribed by https://otter.ai
