Miles Grimshaw - The DNA of Software Companies - [Invest Like the Best, EP.312]

My guest today is Miles Grimshaw. Miles is in his early thirties and is a General Partner at Benchmark. His experience and success belie his age. He was an early investor in Segment, Benchling, and Airtable, all before they had 30 employees. I have learned a ton from Miles about software investing and that’s why I was excited to have him on the show. We discuss his biological approach to investing, whether pure API companies can be good businesses, and what most has his attention right now. Please enjoy this conversation with Miles Grimshaw. Listen to Founders podcast Founders Episode #136 A Success Story: Estee Lauder Invest Like the Best with David Senra: Passion & Pain For the full show notes, transcript, and links to mentioned content, check out the episode page here. ----- This episode is brought to you by Tegus. Tegus streamlines the investment research process so you can get up to speed and find answers to critical questions on companies faster and more efficiently. The Tegus platform surfaces the hard-to-get qualitative insights, gives instant access to critical public financial data through BamSEC, and helps you set up customized expert calls. It’s all done on a single, modern Saas platform that offers 360-degree insight into any public or private company. I’ve been so impressed by the platform that my firm, Positive Sum, recently made an investment in Tegus. We did so because we feel that Tegus will be the gold standard platform for investing research for decades to come. As a listener, you can take Tegus for a free test drive by visiting tegus.co/patrick. ----- Invest Like the Best is a property of Colossus, LLC. For more episodes of Invest Like the Best, visit joincolossus.com/episodes. Past guests include Tobi Lutke, Kevin Systrom, Mike Krieger, John Collison, Kat Cole, Marc Andreessen, Matthew Ball, Bill Gurley, Anu Hariharan, Ben Thompson, and many more. Stay up to date on all our podcasts by signing up to Colossus Weekly, our quick dive every Sunday highlighting the top business and investing concepts from our podcasts and the best of what we read that week. Sign up here. Follow us on Twitter: @patrick_oshag | @JoinColossus Show Notes [00:02:57] - [First question] - His notion of the investor as a biologist or a physicist [00:05:24] - Why he seeks out new companies with unique business models [00:07:53] - How his investments are based on present and future needs in the market [00:11:55] - Evaluating the genetics of a nascent or small company [00:13:38] - The half-life of information as it flows through a company or platform [00:17:26] - Unpacking how software companies can survive re-evaluation periods [00:21:03] - The power of environment creation and facilitation [00:25:10] - The importance of user conferences [00:25:45] - A company’s potential for a differentiated second act as a sign of good genes [00:30:21] - Product quality, timing, and reinvention in tech startups [00:33:10] - Why it’s crucial for companies to avoid copying their heroes [00:37:41] - Breaking down market perspective on pure API companies [00:41:29] - His views on software targeted to vertical versus horizontal markets [00:44:29] - Carefully leveraging relationships with core customers [00:48:06] - Operational lessons from his experience with the companies he’s invested in [00:50:26] - His maxim that software development is as much an art as a science [00:51:12] - His idea of a product magician in the software industry [00:52:19] - Effects of new products and categories at the forefront of the space [00:58:21] - How software founders should prepare for 2023 [01:01:41] - How both market structure and product shape the genetics of a business [01:04:32] - The challenge of pricing and packaging for SaaS companies [01:06:42] - Cardinal sins in software investing [01:07:42] - The kindest thing anyone has ever done for him

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Patrick O'Shaughnessy?

  I know firsthand how complex the tech stack is for asset management firms. And seemingly every new tool and data source makes the problem even worse, adding more complexity, more headcount and more risk. RidgeLine offers a better way forward. One unified platform that automates away the complexity across portfolio accounting, reconciliation, reporting, trading, compliance and more, all at scale. RidgeLine is revolutionizing investment management, helping ambitious firms scale faster, operate smarter and stay ahead of the curve. See what RidgeLine can unlock for your firm. Schedule a demo at ridgeline.ai. This episode is brought to you by Tegus. Over the years of our partnership with Tegus, they have evolved from a pure expert network into a full company intelligence platform. I've been so impressed by the platform that my firm, Positive Sum, recently made an investment

• 0:45Patrick O'Shaughnessy?

  in Tegus. We did so because we feel that Tegus will be the gold standard platform for investing research for decades to come. Tegus streamlines the investment research process so you can get up to speed and find answers to critical questions on companies faster and more efficiently. The Tegus platform surfaces the hard to get qualitative insights, gives instant access to critical public financial data through BAM SEC and helps you set up customized expert calls. It's all done on a single, modern SaaS platform that offers 360 degree insight into any public or private company. As a listener, you can take Tegus for a free test drive by visiting tegus.co/patrick.

• 1:25Patrick O'Shaughnessy?

  Hello and welcome, everyone. I'm Patrick O'Shaughnessy and this is Invest Like the Best. This show is an open-ended exploration of markets, ideas, stories and strategies that will help you better invest both your time and your money. Invest Like the Best is part of the Colossus family of podcasts and you can access all our podcasts, including edited transcripts, show notes and other resources to keep learning at joincolossus.com.

• 1:50Speaker 01

  Patrick O'Shaughnessy is the CEO and founding partner of Positive Sum and the CEO of O'Shaughnessy Asset Management. All opinions expressed by Patrick and podcast guests are solely their own opinions and do not reflect the opinion of Positive Sum or O'Shaughnessy Asset Management. This podcast is for informational purposes only and should not be relied upon as a basis for investment decisions. Clients of Positive Sum or O'Shaughnessy Asset Management may maintain positions in the securities discussed in this podcast.

• 2:22Patrick O'Shaughnessy?

  My guest today is Miles Grimshaw. Miles is in his early 30s and is a general partner at Benchmark. His experience and success belie his age. He was an early investor in Segment, Benchling, and Airtable, all before they had 30 employees. I have learned a ton from Miles about software investing, and that's why I was excited to have him on the show. We discuss his biological approach to investing, whether pure API companies can be good businesses, and what most has his attention right now. Before we transition to the episode, I want to highlight the Founders podcast, which is part of our Colossus network. David Senra, who hosts Founders, has devoted his life to learning from history's greatest entrepreneurs, and every week, he distills the lessons of a different founder. If you want an entry point, I highly recommend starting with episode

• 3:06Patrick O'Shaughnessy?

  136 on Estee Lauder. I hosted David on Invest Like the Best this summer, and it's hard not to walk away insanely energized after listening to any episode with him. You can find a link to Founders and that episode in the show notes of this conversation, and you can search all past transcripts on our website, joincolossus.com. Please enjoy this conversation with Miles Grimshaw. All right, Miles, so a great place to begin is with this notion of the investor as biologist versus the investor as physicist. And the reason I start here is because when you look at the successful investments that you've made so far across your career, they sort of can't be categorized. They're across all the major categories of our style of investing.

• 3:50Patrick O'Shaughnessy?

  And so I think this opening frame will help people understand where you're coming from as an investor and sort of what you've done so far, before we start to dig in on the specifics of some of the companies and some of the bigger ideas that you formulated so far in your career. So start us with this idea of biologist versus physicist.

• 4:06Speaker 02

  I don't know that I picked the easy way to go about it. In some sense, I think the easier way is that of a physicist. And the lens I use there is the idea of having to try and divine a set of rules, a set of axioms that are unbreakable, that you follow, that you look for repetition of, that are consistent and applicable. That's the physicist. You can imagine that these days is the SaaS playbook, is SaaS metrics, is the comparables of, okay, if you got from one to five really fast with this efficiency, it's all going to work. And I personally don't think like that. I think of myself and the exploration, what we get to do in early stage, more is that of a biologist. You might imagine Darwin getting

• 4:51Speaker 02

  off the HMS Beagle, exploring new lands and just amazed at the new species and wondering where they fit into the world and how they mutated to have these adaptations to this environment. And if you think of constantly changing environment, it sort of begets looking for constant adaptations and evolutions and starting with a lens of why, of curiosity, of imagination. At the end of the day, I also think the best companies don't want to be versions of what has been. They want to be the best versions of themselves. And the journey we can go on together is to then be asking, what is our best version of ourselves?

• 5:32Speaker 02

  YouTube didn't want to be the Flickr for video. Amazon wasn't just Barnes and Nobles online. Shopify wasn't Demandware for the mid-market. I think of it as going and trying to be change-seeking. People think of early stages maybe being risk-seeking. The lens of the biologist is one of being change-seeking, I think.

• 5:57Patrick O'Shaughnessy?

  What does that mean in terms of how you show up to a new engagement with a new company? Because it does feel as though the world of software and technology has matured, where there just is more of that playbook-type stuff. There's enough data points now, there's enough of these companies that are building whatever it is, vertical market software or something else, where it's like, oh, okay, we can go do a, how does this work versus the 100 others that have gone before it? And even some of the businesses that you've invested in are probably in that category now, like they're reference points for future entrepreneurs. How does it feel different to you when you've discovered, let's say, a team or a product that has the potential to be one of these very unique, something

• 6:39Patrick O'Shaughnessy?

  of its own versus the next best version of this thing we've seen before?

• 6:44Speaker 02

  Yeah, I think coming at it with a sense of wonder and curiosity and amazement and asking, what's working? What's special here versus how do I fit this in a mold? Take a fun example with Benchling. When I first met them, the classic lens might have been to say, well, it's Veeva but for R and D. Veeva is a fantastic business. I think one of the top performing public SaaS companies these days, a vertically dominant one and certainly one to look up to and admire and be like. But Benchling is on the R and D side, which is very different to being on the commercial side. At the time, it was five people, one piece of software, a single player

• 7:26Speaker 02

  tool. And so you had to dream and imagine what could be. But what was working a bit at the time was some academics were using it, some labs were using. They originally started the company out of a lab around the Broad Institute at MIT because the two founders, Asha and Saji, had been working around the Broad Institute. And so they built it in part for themselves. So they had academics using it. You might have said, well, that's not where the business is going to be. The business is going to be selling to the top of pharma. So ignore that. But if you come at it with the curiosity and the interest in saying, what's the best version of itself? I think you don't try and apply a Veeva template to it. You instead say, if you're going to do professionally, you're going to have done it in academia first. And you start to think about, how do you nurture that? And how do you amplify that? You start to realize that R and D has a

• 8:10Speaker 02

  wide spectrum of small companies that are academic spinouts, small teams exploring, then eventually are bought up by bigger companies. And you start to have a motion, you think about a motion, imagine a motion that's designed for a broader N of accounts, not just the top of the market. For example, I think Veeva, when it went public, had about 150 accounts. Benchling already has 1,000. If you'd applied the template, I think you'd have suffocated the full breadth of potential as opposed to asking, Benchling is its own thing. What can that be like?

• 8:43Patrick O'Shaughnessy?

  We think a lot about two pillars of this process being the founders doing some interesting life's work and the market signal. And very often you find that markets are just smarter than everything else, than anyone else, that sometimes one person can have an insight and get things going. But market or consumer signals, we were just talking about like the insane signup consumer signal around some of these AI tools today. It's just staggering to see frankly, just mediocre tools get hundreds of thousands of signups. We always try to really respect that customer signal, even if we don't understand it. How do you, in your process, balance those two things? Somebody that has a compelling vision for how things should be versus clear data that even if

• 9:27Patrick O'Shaughnessy?

  it's just those early users, the academic users of Benchling or something, like there's some strong signal, even if it's very early on.

• 9:34Speaker 02

  I think there is probably some cases where you can really divine and have a view as to a perfect premonition, if you will, as to ability to chart that 10-year journey as a founder. And you really had that insight. Most of the journeys I've been on have been a little more, there's a problem area we're working on and we get on the field and evolve. And it's less about the premonition, more what's the direction. And early on, I think, asking ourselves what can be special about ourselves, combined with where do we want to get that feedback loops in the market. So I do think early on about how are we training muscle, how are we positioning our team, the

• 10:16Speaker 02

  company, for better odds at excellence, better odds at adaptations to the change that's going to come. An example I sometimes think of is Stripe, early on, for example, didn't go and get Bed Bath & Beyond or Best Buy as customers. They got Lyft and they got Instacart and they got DoorDash and those customers were the market leading ones. They were pushing the future forward and they were going to pull Stripe along with them and Stripe would have to keep up and adapt its needs. At Benchling, we got Regeneron, which is one of the leading COVID antibody therapeutics and a bunch of others. And they were pushing what was possible. Genetic engineering and would pull our needs along with them.

• 10:58Speaker 02

  So I think about it as some sense of a vector to go work on and then loops to help evolve excellence along the way. And then how do you say yes or no to investment? That's a hard element within that. This is maybe where something you and I have chatted about comes into play a bit, which is this hard to define element. But I think about which is what is the business genetics underneath that? Are we going to be fighting a hard fight or sort of a business model trying to capture value that we're creating? Or is that going to likely be able to flow and not be another innovation that we have to have another challenge along the way? And so will we have naturally, hopefully good business genetics?

• 11:40Speaker 02

  And when you spend time with these teams early on, two people, five people at sort of the napkin math. I saw the fun adage of like tigers and house cats look the same as kittens. How do you discern the difference? There's an element of the founder's authentic drive and motivation, the bit of what you're saying, the vision and going to go there. There's a, I think, a sense of those feedback loops you're creating on the playing field. And then I think this question of what are the business genetics and are those predisposed to being good? Because if I want to be a basketball player, I want to have a great equity value. I'd hope that I have tall jeans. Maybe I can pull it off if I'm short and I don't know enough about sports to know who were the athletes that really pulled it off on the tail of the bell curve.

• 12:22Speaker 02

  But I'd like to think we have on the distribution tall jeans versus short jeans for that success.

• 12:28Patrick O'Shaughnessy?

  Maybe explain that genetics concept a little bit more in terms of like practical terms. So if you've got this idea of good and bad genetics, what does that mean when a company is so young? You can imagine evaluating a company's genetics when it's a hundred people or something and you can start to pull apart elements of it that seem reliable and baked in. But at five people, at two people, how can that concept work for you? It seems like it would be very hard to apply at that stage.

• 12:55Speaker 02

  Again, I wish I had more science than artistry to it. So in some sense, it's a bit of a feel. Some things when you get bigger that you think about that we'd all say is great. We were chatting earlier about your podcast with Will Thondike. It's like, what's great? Really low gross churn. That's pretty good genes. And some products you can think about early on as probably lending themselves to that. It's sticky system of record. The joke one might have really early on is just do databases and social networks. You either got real center of mass and just high inertia to you. Put a lot in the database and moving it out is really painful. Or you've got a network effect and it's really hard for someone to leave because the sense of value

• 13:39Speaker 02

  and fulfillment and purpose is there in that experience. Salesforce in their annual Investor Day slides for this year, they had just tucked away in a part of one of the, I don't know, 50 slides was the growth of their fiscal year oh, seven cohort of customers, which I think was something like 47x as of today. And I think the fiscal year 12 was eight or nine X or something. So I need or nine X in 10 years, last 10 years for that cohort. Like, that's great genes.

• 14:11Patrick O'Shaughnessy?

  We could really nail the point home with an example that you and I have talked about before. You've told me before about thinking through the potential half-life of information that could flow through a given system of record, let's say, as something that you could sort of reason about ahead of time, even if the company was very young. Could you maybe talk through that example as one of these genetic markers or something that you actually might be able to reason about very early on?

• 14:35Speaker 02

  So that one is actual database investment, which my partners, Peter and Eric and Chatham, are far wiser on. I think when it comes to sort of application software, which is where I've spent a lot of time between Lattice or Benchling or GitHub and Slack. A lens I learned thinking about this is Benchmark and Thrive had invested in Greenhouse. And I ended up investing in Lattice when I was at Thrive. And they're both HR companies. But I think have very different genes ultimately. So one of those elements, as you were saying, is what's the half-life of the data that's being managed? And recruiting data, the half-life is sort of, if you think about the span of a recruiting cycle, it's probably three months, maybe six months.

• 15:18Speaker 02

  And if you are up for renewal, you're thinking about your software, your application again, what's the switching cost as much if that's true? Lot simpler. I've got three months of stuff to move. Whereas at Lattice, it was everyone's performance management. It was all of the annual and quarterly or half-year feedback cycles. So the half-life, I think, is probably on the order of 12, 24, so. I think a CRM half-life is pretty great in sales. I think about early on, trying to reason a little bit about, it's not a science, but to say, how much inertia, what's the half-life of information that the application will end up managing? There's also sort of how much of an organization you might be able to touch?

• 15:59Speaker 02

  How much do you actually naturally end up with seat potential and cross-functional usage in an organization? Which I think leads to healthier versions of expansion versus coming back to the customer and say, "Well, you grew, but the thing we're doing for you didn't really grow very much, but you're getting more value from it because you're more successful, even though it's still in your department and that it's not really getting seats in there. It's more of an analytical tool." So I think, what's the unit of adoption going to be? And how does that lend itself to sort of better versus harder monetization challenges? If I were to think of three broad strokes to, are we going to have better versus taller genes? I think there's, can we compound inside a customer? Can we grow with them?

• 16:45Speaker 02

  Are we really sticky? Are we going to be able to outlast maybe natural reevaluation moments? Second would be, can we compound externally in the market? Is there an ecosystem that can be built around us? Is there sort of people's jobs or skills mapped on to us that will strengthen around us? Is there an ecosystem to nurture around the company segment? For example, had all the integrations around. The other would be that gets me excited early on is we'll be able to compound layers of product potential. If we succeed here, will we be an interesting place to do a second, third, fourth idea? I sometimes ask myself the question, and it's crazy to try and imagine these, but

• 17:28Speaker 02

  early on, do I think we'll suffer from indigestion? And can you imagine being at the board meeting five years from now, looking at the next couple of year strategies, more of an exploration? And you're like, wow, there's still so much to build. You don't feel like you'll ever be done. I think if you can compound product curves like that, and the first one hopefully sets you up for that. If you look at the very biggest companies, I think they have this attribute, and I think that sort of sets up for, again, interesting genetic material to work with.

• 18:00Patrick O'Shaughnessy?

  There's a couple ideas really within each of those that are interesting. In the first one, you said, can we imagine surviving a period of reevaluation? I'd love to understand a little bit more about how you might think through that specific hurdle. We were joking before that so many people will say something about investing, and what they're really saying is that you want low gross churn, which is definitely just true. And maybe this is just yet another example of saying you want low gross churn and something that will drive that. So maybe the way of asking the question is just like, what helps drive low gross churn, in your opinion, to survive reevaluation periods, to have very low customer turnover? Because that's where all the cohort stacking, like what you referenced with Salesforce, can become so powerful. So in that first genetic

• 18:44Patrick O'Shaughnessy?

  bucket, say a bit more about reevaluation and gross churn that it implies.

• 18:49Speaker 02

  The place where I probably thought about this the most, and I think it's broadly applicable to think about, is in the HR context. An organization will have a different mindset at different stages of scale. At Lattice, we said early on, there was a question of, we started in one product area, would you go up market? Would you go broad? The hard thing about being up market is there's natural graduation moments, if you will, within accounts where you started really small, you didn't have anything. You picked your first set of tools. You worked with those for a while. You got sophisticated. You hired the C suite executive. The C suite executive came in and says, I'm going to level us up now. What do I do? Well, I go buy some new tooling that does the job at the next level for the next set of scale.

• 19:34Speaker 02

  That isn't backwards looking, that's forwards looking. And this executive coming in says, now's a moment for us to level up and look forward. And they come in and pick new software to help with that. I think that's the workday transition when you get really big. And so if we're in an account, you're naturally going to get the question when the C suite executive comes in, CHRO at 2,000 people, the workday transition is going to be the thing to draw for a couple of years. And you're asking, so how am I going to survive that? And instead, I think you could say, well, I don't need to try and figure out how to survive that in some sense. Okay, you're going to eventually graduate off. We're going to build for holding you from your first decision until that transition. And we're going to have a board sweeten that. I think developer products have a bit of a similar moment potentially.

• 20:18Speaker 02

  And something I've looked for and thought about early in partnering with them, and maybe incorrectly have thought this would be a risk and it hasn't applied to some, but is when will that build versus buy moment emerge? And I don't think there's a stark moment when a company sort of finally says, hey, should we actually go build this now? But I think the beauty of the API business is you can get it really early for a very specific job. And the beauty is when that job scales, think of Stripe, think of Twilio, think of Plaid. There are some others that are more integrations that I think there'll be a question of graduation risk over time. Go to hosting sort of application, you had Heroku, which is really great for getting started.

• 21:02Speaker 02

  But I think if you or I were working with a relatively successful software company and they're still running everything on Heroku, we'd be asking some questions. There you get a maturation or graduation risk. And so I think the beauty is when you can imagine it being able to either span that journey or again, you acknowledge that genetic trait and you say, like in Elatus's case, we're going to go broad and just expect it at some point. We're going to hold you quite a long time. It takes a while to go from 50 to 2,000 people. We'll hold you for that period versus maybe fight against it with wasted energy.

• 21:36Patrick O'Shaughnessy?

  Obviously, like, none of these answers to the questions I'm asking are universal. You said at the beginning, there's a lot more art in many cases than science, and all of these are sort of N of one thing. With that being said, in the second bucket of good genetics, you talked about the capacity for, I think of it as like a platform, that the company has an ecosystem being built on it or around it. Are there consultants that get hired because they're good at X, Y, and Z, or something like that? Or, in Segments case, it integrates with every other tool, so you can build all these workflows on top of a segment instance or something in a company. Say a bit more about when you feel that second bucket of genetics, for what kinds of companies

• 22:17Patrick O'Shaughnessy?

  that second type of good genetics, like platform potential or ecosystem potential, is more or less important? Because I'm sure in some businesses, it's really key, maybe in vertical market SaaS or something, it's less applicable or something like that. So where is that notion of the integration potential and the ecosystem potential more and less important in your mind?

• 22:36Speaker 02

  In that sense, it might not be possible, but it's something one universally, I think, should be vigilant to and be thinking about. Trying to ask the question of, what could the best version of that be for us? There's the classical elements of it, that people built stuff on top of us, and that's beautiful. People build businesses on top of us, and they're only strictly empowered by us, and we're taking a smaller piece than all of them, and that's very symbiotic platform relationship. There's an element that are lighter that can be applicable in some of the application software contexts that are really interesting to think about. Some of those is someone had a job and they brought you in to help them in their company. In some sense, there's a sort of meta-game that's getting played around your application,

• 23:19Speaker 02

  which is their professional success and their power and status and sense of excellence and sense of importance inside an organization. What at Figma, do I think in many ways it was really interesting? Yes, they sped up collaboration, but they also made it so that everyone could actually really see all the designers and all their work they were doing. They really elevated design inside of companies. Because instead of being a Dropbox file that you didn't know which version it was, and you tried to open it and you didn't have the specific version of the software to see it, you can now just go to a website and the copyright could be there. And so they sped up collaboration, but they also elevated, I think, design that org. And there's this question of how are you also helping the career needs, reputationally

• 24:02Speaker 02

  is giving superpowers to your end buyer, your end adopter, your end user inside a company? And that sort of meta game, in some sense, that ecosystem game is happening around. How could it be that someone by adopting you gets that promotion? Okay, it's not platform, obviously, but it is this compounding element in the market external to you that's interesting. Can you facilitate it? I think there's also this lens in some SaaS that I've thought about, which is you have what urban economists called the agglomeration effect. And in thinking about urban centers, there's this curiosity, and economists always have to make a fancy term for it. But what happens and you end up seeing as emergent phenomenon is all the bakers are

• 24:46Speaker 02

  on the same street, all the butchers are on the same street. Or why do you have Silicon Valley as a really dense aggregation point for a lot of software and silicon? It wasn't preordained. No one said, okay, you've all got to come here. Urban economists call that the agglomeration effect. And instead of saying, well, the bakers are on the same street, shouldn't the bakers be dispersed throughout the city? Wouldn't that be better? That computer will sort of know if they're all there, everyone aggregates that. And they actually all succeed better by being next to each other than all across it because you've dispersed demand. And I think there's a potential, it's not a network effect on the company, but this question of how can you sort of have an agglomeration effect for your application software? How is it that the marginal integration potentially with an ecosystem partner, a

• 25:29Speaker 02

  marginal executive in the market feels they got a career fast-tracked because of using you? How is it that acquisition channels feel most predisposed to sort of you and they will succeed because you're succeeding?

• 25:43Patrick O'Shaughnessy?

  What do you think about user conferences in that context? To nail home the agglomeration effect, why does the idea of a user conference seem powerful to you?

• 25:54Speaker 02

  The celebration of what is possible, individuals sharing their progress, how they've helped their organizations, how they've solved problems, maybe how they've gotten a career boost, even just being proud. A user conference is in some sense, setting up the street and inviting all the bakers to be it because you got all the best critics to show up. And trying to think about being that is a worthy endeavor.

• 26:21Patrick O'Shaughnessy?

  And the third category of good genetics, you talked about the potential for a marginal act to the company, a second product, the third product, whatever. Segment was, we've talked before about why it's such an interesting example, because its second act wasn't even selling to the same original customer or buyer. Seems like unusual. Often you'll hear people say you want to continually compound value for the original customer, which is kind of part of what you said in the first bucket of genetics too. But talk about that capacity for a second product, which seems like an unpopular thing these days. The narrative is focus, focus, focus, just do one thing, be the best at one thing for a really long period of time. There's more nuance to it than that, probably.

• 27:02Patrick O'Shaughnessy?

  How do you think about the capacity for and the strategic decision to go after a second act?

• 27:07Speaker 02

  Again, there's no one rule to rule them all. But I think if one looks, and it's certainly been my case with many of the teams I've had the fortune of partnering with, that it's happened earlier than you might imagine. And there is this adage of focus, focus, focus. Now, you don't want to do silly things. You also shouldn't assume success of the second thing and over resource it and build it in the dark for a long time. But I think the muscles that you end up building by asking yourself where else we go, what else we can do, and starting that curiosity early, earlier than you might imagine, is powerful because it's not going to happen instantaneously.

• 27:48Speaker 02

  And just starting to ask the question begets engaging more deeply, I think, with the market. Yes, you could do the ivory tower version of we're going to go reinvent it. That hasn't been the case of the teams I've worked with. It's actually more encouraged being in the market, being close to customers. Customers and asking what other jobs are there to be done. Where can we go with the beachhead we have? I looked at Viva, which I think actually started working on Vault. I think around like 50 or 70 million of revenue, give or take, certainly south of 100. HubSpot was working on the sales product. It was just about launched when they went public. I think it was about 70 or 100.

• 28:29Speaker 02

  So maybe it was the year they crossed 100 where it was staffed, team working on it. Started with a small acquisition, actually. At Segment, we were the one integration for every integration for all of the customer data. Peter started and the team started asking early on, what does connecting all of that enable? Well, it should enable you to be more customer centric. It should enable you to have a more unified view of your customer, even though you're using discrete best in class tools. What is that going to look like? How can we enable that? That led to, we started a team focused on that, probably about 30 of A R. By six or 70, we launched it as personas at a big user conference with a bunch of partners, ecosystem partners.

• 29:11Speaker 02

  And that was an interesting learning as well because it was sold to the marketing team, whereas predominantly we'd been landing with engineering. We were an API business. So engineering had pulled us in and this was to create value for the marketing team. There's some fun lessons around not copying your idols. And at Lattice, speaking of genetics, HR is a hard market, highly competitive. And in that case, we were third in the market, reflective, which actually ended up selling for like five million bucks was way ahead of us. Raised $100 million, cold ramp was out there. We were third. And the view was that we're going to bundle or probably be killed. And so existentially, we should build a suite and use that to differentiate.

• 29:53Speaker 02

  And so I think at about five or seven million of A R. Jack made the bold call and took basically all of our EPD resources off of performance management and put them all working on an engagement product, which we launched by about thirteen or fourteen. It was shitty, but it worked. And then we were the only company selling into the mid-market that had both. And close rates went way up, sales pipe went way up because we were suddenly differentiated and our competitors who were bigger had to react by acquiring people and then dealing with integration questions and everything else. So you go look at HubSpot now or Datadog or Salesforce and the number

• 30:34Speaker 02

  of clouds they have, being able to layer in these customers, obviously over time is incredibly powerful. If you want to still be compounding it, many hundreds of millions of revenue, be reaching a billion. And I think it's a muscle that if you wanted to probably be there at a couple hundred million working, probably needs to start earlier than you might realize.

• 30:55Patrick O'Shaughnessy?

  You said two things there. I want to definitely come back to the not copying your heroes because that's just appealing on face value, that phrase. But the idea that you launched something that maybe was shitty, but it worked. I'm really curious how you think about product quality versus timing? Because the answer there seems to be like, well, I need something that does both of these things. And if both kind of work, the bundle, I guess, is more valuable to me than two perfect point solutions or something. What have you learned about product quality and when it's more or less important? Because there's plenty of examples of enormous companies where you could probably say like the product sucks, but it still does the job and it's really, really sticky. Product quality does

• 31:39Patrick O'Shaughnessy?

  not guarantee great business outcomes. So what have you learned there? Because that seems like a really key thing to get right.

• 31:46Speaker 02

  I tend to be in the camp of earlier the better. At a Stripe, you probably don't want to launch an incredibly shitty version of something that's meant to have five nines of performance and be used to move your money. And when it's down, it doesn't work, you can't make money, or the Instacart driver can't buy the groceries at the store when they're waiting to check out. There's, I think, some categories where probably it's a little more existential. Product quality is part and parcel with working upon some amount of success. Let's assume you're working on a second thing, but the first thing was really successful. There can be a view of, well, we can't launch an MVP again. We can't have sort of a beta moment. Our customers won't be that forgiving with us.

• 32:29Speaker 02

  And part of that reinvention muscle is how do you not let the bigger company operating cadence and structures that are going to get built and rules not be the only way of being, snuffing out. And I actually think Patrick at Stripe has done it because they've kept reinventing quite a bit. They have an operating cadence of there's quarterly business reviews and then there's some high tempo reviews. It's way faster. It's a smaller product. And the key question is, how can we unblock you? How can you move faster in those versus more of a quarterly cadence? So not letting perfection, it has to be for everyone, big company annual or quarterly

• 33:09Speaker 02

  cadence, suffocate the muscle of going zero to one again inside of it. If you actually go look at some of the bigger companies, I think HubSpot does this, they'll actually just end up launching and then relaunching the same product. Here, here's an alpha version and they'll work with it and it'll be a market for you. They'll add some more features to it, refine it even more. And then they'll say, now it's available to everyone potentially. And it's almost a relaunch moment. So I think one is better served not stifling and having perfection from scale and from success be the enemy of innovation.

• 33:44Patrick O'Shaughnessy?

  You mentioned this idea of not copying your heroes and how that can be really important. You, you mentioned a little bit of this earlier. Don't just be the best version of something that came before you. Expand on that specifically because heroes implies like a person versus a company. So why not copy heroes, which so many people build their whole careers effectively copying the greats that have gone before. So it's a very common thing, which makes me intrigued by the notion. So expand on that notion.

• 34:10Speaker 02

  Well, I was thinking about it in the business context. Versus a person context. One place where I thought about this a bunch is at Segment, we were an API company. Sold an API in many ways, we sold to developers that got pulled off the shelf early on. It was open source in many ways, analytics JS, and got pulled off the shelf by engineers. So we were quickly, hey, this is Twilio, this is Stripe. And that created a mindset, because those succeed in this way of, oh, if we just have good docs and engineers can just use it and they'll put it in and once we've sold it, it's theirs and we should price on an API call basis because that's how everyone prices on an API call basis. And the reality was, yes, engineering adopted it.

• 34:55Speaker 02

  And there were some early-stage teams, some forward-thinking engineers who did pull it off and just use it. But for a lot of the customers, there was real change management needed. This wasn't, oh, just put this in and text messaging or run through it. This was, which order do I integrate everything? I already have integrations. Do I do it all in one? Do I do it piecemeal? Can I do a piecemeal? Redshift was just emerging in some ways at the time and people were starting to want to dump it into Redshift. Can I do that at once? Is that actually there yet? If it's not there yet, should I wait for everything else? My marketing team now touches a bunch of these integrations. Is that's going to mess up them? Should they think about it differently? So there's real change management. It was almost maybe more of like a database migration than a, here's an API call, you just use it.

• 35:38Speaker 02

  Internal change management, some amount of cross-functional education. We'd had a mindset of, okay, we sold it, you bought it, we can answer your questions. As opposed to working through those complexities is part of excellence, part of success in the product and we'll be best at that and we'll take ownership and accountability for that. So we actually, later than we should have, in part because again, we copied the API companies and had a lot of that influence, went and looked a lot more like the vertical SaaS companies that actually do do implementations, have professional services, have packaging of support and activation, have programs and best practices for rollout and built out a decently large PS organization.

• 36:21Speaker 02

  And we went from a bunch of accounts, you'd be sort of amazed, I can probably now share it because it's well past its history books, but we had some million-dollar deals that by like month 12 had never been implemented. They were just shelfware. They were sort of bought because of, it was future-leaning, we created this category, the CDP. And if you went and looked at some of these accounts, a big portion of revenue, nothing was really piping through. Or if they had put something in, it was like two things on the side and it didn't really matter. And we started packaging up PS, we set expectations of rollout, we set excellence for that, we figured that and that radically changed. And the other element again was copying your idols. You would have said, well, let's just do API based pricing. That's what everyone does.

• 37:04Speaker 02

  Well, why do we want you to decide which APIs to use or not use? Is your analytics data different than your email data? What's the value of each of those? And the thesis was what we're really aspiring to is we're going to connect up all of your customer data. You can use best in class tooling, but you can ultimately get to one unified customer profile. And so at about 30, I think or so, we controversially at the time, but again, asking what is our purpose? How should we best realize that potential versus do the standard? Switched to what we called MTU pricing, monthly tracked users. The theory being we want you to put users with us and we want you to give us as much as give us everything. Don't think about it all.

• 37:50Speaker 02

  How can you make pricing and packaging an amplifier of what the products intentional, what the products genetics are versus a potential detractor inhibitor, et cetera? Those are two stories I think about is not taking the prior templates. And in many ways, looking like Twilio and Stripe is fantastic. And so it was very easy to just say, we're that and not inspect ourselves.

• 38:14Patrick O'Shaughnessy?

  I'm always interested in taking the market's realities and sussing out what some of the things it's telling us might mean. And one of those things today is everyone talks about picks and shovels and technology, or that's like a category of business that just in general, people like. And APIs are the ultimate pick and shovel for software. But if you look at the market, the market seems to think that database companies where there's huge market caps and large multiples, also kind of a pick and shovel are just very different and more valuable than pure API companies. If you take Twilio, for example, today, like its market cap is low, its multiple is tiny. I saw somewhere the other day that it's trading at 33 times its interest income or something like that. So can pure API businesses be good businesses?

• 38:59Patrick O'Shaughnessy?

  The market seems very skeptical. Of that potential, whereas it's always blessed database companies or companies with network effects or systems of record, which are just like database companies. Can there be a good pure API business, do you think?

• 39:14Speaker 02

  I think so. There's two factors. One is you hear people talk of, is it a commodity? Because 50% gross margins, the whole business, what's actual messaging like, et cetera. I don't ascribe not through like some scientific rationale or I've really inspected how carrier relationships work in all jurisdictions. And so I know it to be the case that that isn't real commodity. But sort of the stickiness of the customer relationship seems pretty powerful to our point. They're at four billion of revenue and NDR was still 120%.

• 39:52Speaker 02

  To have that level of historical cohort growth at that scale implies to me at least really sticky. Now, that's not to say it wouldn't be sort of commoditizable, but if it was really commoditizable, I think you'd see people moving in and out, or you see so much pricing pressure that even growth within the accounts would mean you probably didn't see that sort of an NDR. I think the flip side, though, is it's confounding in that and looking at that one and specifically is they spend a lot of money in part because they're working on lots of new initiatives. And there's a question of how successful those new initiatives will be. But you have a business that's four billion of revenue, two billion of gross profit and well, free cash flow negative still. And if you actually believe stock-based comp is a cash expense, some would say that's

• 40:35Speaker 02

  an opinion and dilution's the only factual element, whatever. It spends a lot of money. Does it structurally need to do that? Not clear to me that that would be the case. So as a shareholder, you're aligning with leadership that the scale of investment is going to yield really amazing returns on new initiatives. And they are working a lot of new initiatives, everything from segment to flax to video, et cetera. But as a core business, if you were just able to sort of pull it out and look at it. I don't think we'd say that was a bad business, but we'll fun to see Stripe eventually go out and get to look at that. And the big question to me on a bunch of these over time

• 41:15Speaker 02

  has always been graduation risk. Can you get these customers small and really have them stay with you? So this build versus buy question at scale, the only name in a Twilio case that I think we all know of as having come off is Uber, which obviously put the stock in a bad place for a while when it did. But I don't hear of many others saying, I'm going to do a build versus buy. It might be alternative buys and value of the platform and other things.

• 41:42Patrick O'Shaughnessy?

  It's interesting to think about. Most of these are very simple tasks, the API, what it's performing. And like, the more complex the guts, probably the better in the builder spy equation. Stripe's thing is really complicated, especially as you go international and all the handling that it's doing to do the same simple core action is insane. Whereas maybe Twilio's is like sending a text is like, I'm sure there's lots of complication, but maybe less than payment standards around the globe or something like that. So complex guts could be something to look for. I'm curious, in that same vein, if you're thinking about some of the other categories, the three others are vertical market software, horizontal market software and consumer, which we really haven't talked much about yet. If you have similar ideas or conceptions or like spikes that you care

• 42:27Patrick O'Shaughnessy?

  about in those key areas, maybe starting with vertical market software, you said something to me one time, which really stuck with me, which was you want in your salespeople, good breath only. No breath is better than bad breath and the importance of sales specifically in vertical markets. Maybe we could start there. What have you learned about vertical markets that are specific to that style of business?

• 42:48Speaker 02

  I think in vertical, you know exactly who your customer is. The sort of idea of working backwards from the customer, of solving for the customer, of that customer connectivity, I think is ingrained from the get go. If you do it properly, if you do it right, you'd rather sort of not show up than show up poorly. You don't probably get many redos with accounts. Who are we building for? What does it take to make them successful? You really get a tight feedback loop on that. I think with horizontal, there's a lot of pressure on the more you allow that to just run rabbit, quickly it will become really chaotic frenetic energy as opposed to in a vertical, we can't get that frenetic. It's fairly constrained.

• 43:31Speaker 02

  So in a horizontal, I think you have to very early on start thinking about what's the systems? What is the repeatability of the motion? How do we segment the market? Where do we really want to be sort of winning with repeatably? And with what messaging versus where's more experimental? And so the sort of go to market system design question is brought to the forefront earlier with consumer. It's somewhat hard to motivate behavior, but if you motivate it, you're always thinking about what's the network effect? What's the behavior we're tapping into and fulfilling? What's the behavioral experience we want to give? How do we make a, depending on the product, obviously, but how do we maybe make a champion of them? The Musically CEO had this great lens. Musically was ultimately bought by TikTok.

• 44:16Speaker 02

  Sort of thought about it as creating a new land, a new country to go populate. And how do you incent people to come to this new country to populate? How do you create an upper class eventually of them and make them successful, make them stars of the new land so they inspire others to come to it? In a consumer world, you sort of train that mindset early on of what's maybe the atomic unit of the product and how's the network effect, the behaviors get a compound around that in success. I think those are the spikes of the mindset and how could you maybe use some of them in others? Compounding around a core behavior, I think is an interesting lens to think around in a Figma context. The design asset and how other behaviors, other parts of the org have internal network effects and you can unlock those.

• 45:02Patrick O'Shaughnessy?

  Your partner, Chathin, was incredibly influential in my career, building software around his notion of design partners, where you pick five to 10 early customers that effectively determine a lot of what the product becomes. So therefore, picking those customers well is a huge part of like what's going to happen in your business. How do you advise people on customer selection? You said earlier, let some of these great examples hitch themselves to companies that were themselves moving really fast and they were just trying to keep up. So they were sort of getting pulled rather than pushing stuff down onto companies. When you talk to founders about customer selection in terms of what to look for, but also what to avoid, how do you advise them?

• 45:46Speaker 02

  I think at first is thinking about it and caring about it. And again, in this physics version of investing and building a company, did we go one to five at some rate with some burn efficiency and some cash burn multiple? And if so, all is equal. And I think in 2021, there was a lot of, if it's 20 of ARR and growing more than 50%, it just works. These questions don't matter. I think we sort of come back from that view of the world and excellence is separating from good. So who are the customers? I will often spend time with the team early on. What do we hope to have learned? Who do we hope to be serving? And what feedback loops could we be setting up there at 5 million of revenue?

• 46:29Speaker 02

  Where I might say, we'd rather have three versus five of revenue because that's stronger and it's going to set us up better for success. I think each market has its own nuance and it's the founder's insight to determine who are really the great customers who are pushing. And I think that's where great judgment there really separates. I think there are some traps. I think who's big and shiny, but going to be really slow and you're going to sink a ton of time into and it's not really going to be work. Who's a little more backwards leaning potentially versus forward pushing in the market? Even if they won't sink a ton of time in, they're not a terribly innovative organization themselves. And so how much are they really pushing even if they're a decent?

• 47:10Speaker 02

  So I think that's where a founder's special judgment really does exist. But I think asking who's pushing the market forward? And going to really need us, we can really unlock them. Think of Robin Hood for Plaid. Think of Uber for Twilio. Think of Shopify for Stripe. Think of Regeneron for Benchlink. All of these, I think, had some of them in some way. You take Segment. The whole point was to connect up all your customer data. Therefore, everything should run through us. Really hard to run all of that at scale. One of our biggest and most challenging customers early on was actually Hotstar, which is the cricket show of India, which literally 10x is cricket shows. And so I would forget what day of the week it would happen, but when it would happen, the whole platform's volume was

• 47:56Speaker 02

  Ten X every other day as a function of them, really bad for margins. We talked about graduation risk. There's also bad customer risk. Almost every team I know I've worked with has let go of a customer. Segment let go of hot start, Benchling, we did a, for research, you actually want to synthesize DNA, actually like e-commerce order DNA. And we powered a website for that because it was kind of interesting early on and company called Gen Nine. And we eventually let them go at like a million of AR. So I invested, I don't know, 700 K of AR and I went back to 500 K. Lattice, we had a big customer who was a couple thousand seats because we're like, oh, maybe we will go up market. We let them go. We weren't built for that. So there's good customers and there's actually knowing who are bad customers and letting go.

• 48:39Patrick O'Shaughnessy?

  Any other operational lessons that you feel really sing across your experience? I mean, it's obvious from this conversation and from others that we've had that you really have a sense for what's going on in the companies that you're investing in. You're really partnered with them. Are there any other big things that just scream out from your experience on the operational side? That we haven't talked about yet that you think are really potent and powerful?

• 49:05Speaker 02

  So I think one commonality is that question of good or bad customers early on. Who are you designing those feedback loops? I think that's really worth being purposeful about early. I think another one, which is sort of a consumer learning in some sense, applied into B two B SaaS. I think in many different formats is being rigorous early on about product metrics. In a B two B context, sales metrics can rule the day. And the bit to the segment story, we had some accounts that weren't turned on. And then also trying to say, what's full realization of potential is everything's connected.

• 49:45Speaker 02

  So we should be able to see in our customer data that we have a whole bunch of tools in terms of categories of tools that are used. And if we don't, we should think of that as a failure, even if we have the revenue. I think about sort of a fun analogy I occasionally use is TikTok versus Instagram, at least Instagram as of maybe now, maybe six months ago. This idea Stated versus reveal preferences. Instagram is your stated preferences, who you wanted to follow. That's who fills your feed outside of ads. TikTok is really your reveal preferences. It's the videos that actually give you the dopamine hit and you really want to watch, which may or may not overlap with the Instagram ones. And I think of sales metrics and product metrics in part being the yin and the yang.

• 50:29Speaker 02

  The stated preference is what the sales relationship and the customer was saying, I want this. Doesn't necessarily actually mean it is going to get the job done or be adopted or be realized in the way in which you hope it is inside of the organization. I think consumer forces that mindset, obviously early because that's all you have. But I think that thinking in a B two B context of how should our product be used? How do we hope it's used? And measuring against that is a really healthy foundation to put in early.

• 50:59Patrick O'Shaughnessy?

  If you think across all of software investing, where obviously you've spent so many hours, so much of your life, is there one thing you believe in most?

• 51:09Speaker 02

  I do believe software is as much of artistry as science. The elegance of great software, it's sort of the craft of, again, product, I think, sets the genetics of the business inside a customer. And I don't think that's a science. I think there's a lot of artistry and thoughtfulness to that. In the same way, we know that in the consumer context and we think about that. The nudges, the behaviors, the feedback loops. I think a lot of it does exist in the B two B context. I don't think it's pure utilitarianism. There's an artistry and a beauty to really cracking it well.

• 51:46Patrick O'Shaughnessy?

  Without thinking hard about it, when I say product magician, who pops in your mind?

• 51:51Speaker 02

  I would say Dylan. The beauty of that simplicity of the product, getting it to work really fast in a browser and.

• 52:02Speaker 02

  Without defining it, but think about business genetics. That's really great. You own the design artifact for the core product experience. How many other people touch that and are going to be seats in that? So your depth in that account easily copy over and sort of move everyone. Cause you've got a lot of cross functional work happening. So you both have asset control, data control in that sense. You have cross functional, internal network effects of people around it. You have natural seat monetization that can come from that. You have an external ecosystem that gets built around that. It's a skill that people have. Think about recruiting a designer right now. If they said, I don't know, probably not going to get recruited.

• 52:45Speaker 02

  I'd say you couple great genetics with a simple but beautiful magical product experience.

• 52:53Patrick O'Shaughnessy?

  I'm curious for your take in the current environment on two things. What's going on in product space? Maybe talk a little bit about AI or some of the enabling things that have emerged. And then what founders should be thinking about, because obviously like their world, if you're raising, has changed so dramatically from 12 and 18 months ago. But let's start with product space. What is attracting you right now? Like what most has your attention as important developments in terms of what has now become possible to build great new software products?

• 53:23Speaker 02

  Well, there's AI, but before we go to the tip of everyone's tongue, I think there will still be great opportunities where new market segments are emerging. Benchling, I think, is that and what's happening in sort of genetic engineering and biotech, chain analysis, which my partner Sarah invested in for the crypto market that emerged and money movement on various blockchains. There's always going to be new market segments that emerge. Again, that constant curiosity adaptation is interesting. In some sense, new monetization models around software. I think you see Ramp and Brex. I don't have the data on it. I don't know exactly how well they're doing, but if I just step back from it all and look at

• 54:07Speaker 02

  using interchange revenue simplistically in lieu of.

• 54:12Speaker 02

  Software revenue, classically, it's sort of an interesting twist on monetization model around software. I don't know where that could go, but always curious for lenses there. You might even put toast in an interesting category of a broad suite against the customer segment and where can you use other parts of a suite to monetize or maybe give away more but monetize in other areas? I also don't think we're probably done with adapting work to the new age. I'm actually a big fan of going back in office and more in office versus remote. I think teams, even if they are not remote, will be more distributed earlier on. And so, let's say a hundred people,

• 54:56Speaker 02

  you might have three hubs when previously you would have one. And you're going to have teams working across those hubs. You don't want to have all of your engineers in one, all of your salespeople in another. You'll have mixes. It will feel much more distributed. And so, how can collaboration around work keep getting made? He says, I don't know whether show up, that's the genius of a founder to see that adaptation and evolve it. We talked about Figma. Today, Figma is an interesting adaptation of how to enable better collaboration around design assets. Gong is an interesting one around how to have better collaboration, sales and training and enablement all around the sales call. I actually think it's facilitating better sales, but it's also the collaboration around sales. It's not just the system of record anymore.

• 55:41Speaker 02

  It's not just the artifact of who are we talking to and what stage are they in? It's how do we all work together on an account? How do we get feedback from each other in sales calls when you hear of teams adopting it and using it? There's, I'm sure, more that we'll figure out for that, I think, that new, permanently more distributed world. And then, obviously, I think in AI, what's just showing up viscerally in ChatGPT and DALL-E is just crazy. And where does it go is a really interesting question that everyone's ruminating on and exploring that I think you should absolutely have broad, open-mindedness about.

• 56:22Speaker 02

  As we were talking about the language side of it and what's possible certainly I think makes existing product experiences really interesting. I think most companies we all work with should be thinking about its application to them as well in terms of product capabilities, product experience, et cetera. What avenues does it open up for either a totally reimagined product experience or a reimagined solution to a job maybe that didn't exist? I think in some sense Jasper and the others is somewhat indication of where might have a really great

• 57:02Speaker 02

  fit in some ways, which is it's replacing labor. That's a very different mindset potentially than it's a new system of record or something.

• 57:11Patrick O'Shaughnessy?

  And not media labor. In many cases, high end labor.

• 57:14Speaker 02

  Yeah. I think there's a fun lens around it that it's a lot of jobs that were hard to do, but sort of somewhat easy to evaluate. If I asked you to draw a picture of us in this room, it might take you a while. I don't know how good of a drawer you are, but if you've sent it to me, the output eventually, that's a problem to be able to grade it and feel decently okay about my grade. That's sort of different to what we would have imagined, I think of what was possible. We'd have said robotics would be handled because it's really easy to understand what you have to do. I don't know if I tell you to make a picture or write an essay, did you start with the ending and work backwards? Did you start in the middle? Did you do an outline? Did you do black and white first? I don't know. And I don't have a rulebook and you might figure it out. Whereas if I said, hey, manufacture this thing, you'd

• 57:59Speaker 02

  eventually write a script for it and it would have felt more accessible. And it's interesting how the very complex, but still easy to judge at the end, has been unlocked here. And so where does that go? There's probably some interesting, it's already happening in marketing, obviously. What does that unlock for teams there? Jasper is marketing. Maybe in the past, there's the mobile comparison, which I think is not quite right. There's a new distribution channel. The element that is right is probably a new form factor. I think in this case though, I've so far been impressed with how quickly some of the big incumbents have sort of made their mobile app and are adding these capabilities in. You see Photoshop working on it. And it's early and maybe there's a whole re-architecture that needs to get done.

• 58:43Speaker 02

  And I think there is a point of the whole UI could look different in a AI versus fine-tuned world and will that be in the same product experience? But it's certainly new land to be roving in for evolutions.

• 58:55Patrick O'Shaughnessy?

  What about founders? How should they think about what's going on right now? And especially just the change in prices of equity, like the cost of capital has just clearly gone up on average and there's exceptions and some great companies are still pricing at really, really high prices or high multiples. But what advice would you give founders here uniquely at the end of 2022?

• 59:15Speaker 02

  So I think everyone's doing annual planning or thinking about the next year and many are probably doing cuts. I think there's a lot of good tactical advice around that, whether it's inspecting sort of the ratios of various functions to engineers and quota carrying reps. If you're really trying to grow really fast, those are the two hardest to hire, but really the bottlenecks for any company. And so how are those ratios? How are ratios in specific functions potentially? How's management span of control and sort of rewrite sizing an organization? Asking yourself what's must do versus nice to do? What could you sequence a little more and the like? One area or two questions that I've spent some time on with some of the teams I work with.

• 59:59Speaker 02

  One is where could we maybe change assumptions in how we run the business? And so specifically what might an example be of that? Well, maybe we were really aggressive on customer support. And actually, like, we don't need to be anymore. Maybe we've established stronger market position. It's not where we want to get lazy, but the capital available to the 15th competitor potentially that does exist might be harder to come by. Maybe our support model could be reimagined where maybe as a market segment that we serve, not one that is really great for us. I don't know the history on it or I don't know it's right or wrong, but you saw Brex get out of a market segment. Ignore right or wrong, wise or silly.

• 1:00:42Speaker 02

  That question, I think, is an interesting one to pose. Every team is probably operated with a certain set of assumptions about way of serving the market, segments to serve, models internally around that in terms of go to market motion or product features and capabilities. Where could that be worth reexamining for a changed state of the world? And I think the other one for slightly later stage teams. We've talked about second product areas and expansion. Are we really as confident in sort of the BCG cash cow nature of our core? If we actually start to try and do the back of the envelope of separating our investments for the

• 1:01:25Speaker 02

  future from maintenance of the current, is that maintenance position rock solid? I think it's worth, okay, if you're 20 million of ARLs, but if you're 100 or so and you're still investing aggressively, I think it's worth taking the time to be confident on and maybe to reevaluate some, connect to the first point, reevaluate some of those assumptions on or rethink some of that model, be able to then also be clear about, you might ask the extreme of the question, what's the minimum it takes to like maintain current? You could at least start from there and say, everything above that is investment. Am I getting good return on that investment? Do I feel good about that? It's sort of optional in that sense.

• 1:02:08Speaker 02

  It's an interesting foil of a spend number and position to be thinking about.

• 1:02:15Patrick O'Shaughnessy?

  Whenever I talk to you, I've always told myself writing down company comparisons and I have one final one to ask about, because I don't know where it came from, but I just remember writing it down. At one point I talked to you and I wrote down OnlyFans versus Patreon. Do you have any idea why I wrote that down?

• 1:02:28Speaker 02

  One, because we're probably talking about really interesting emergent phenomena. Two, maybe in the bucket of product shaping different genetics. And maybe the market shapes it as well in this case, but they are different markets. If you categorize the OnlyFans market, as most people would, Patreon doesn't serve that in the same way. And so it's both the market structure of a product that leads into very different sorts of businesses. Both would be put in the bucket of subscription businesses, consumer subscription businesses, because both have a subscribe to someone that you're interested in and want to support phenomena. Patreon does it with a, well, which tier do you want to be in? How much do you want to support them? And it's interesting to think about the emotional setup that creates a sort of evaluating and on entry,

• 1:03:13Speaker 02

  making a judgment call of how much you support them. It's not good or bad, but it's an interesting product manifestations, the behavior. OnlyFans, if you've gone and found someone, is a single, do you subscribe or not? And for how long? So the only question is, what's the duration of your support for that person? But underneath that is heavily architected around messaging and actually very geared towards creators on it, creatively using messaging primitives to charge more and to have more conversations, closer fan relationships. So you see an emergence of, in that, a whale customer, almost more of like a gaming dynamic where you have a subscription business. But in gaming, the question is, who are the whales?

• 1:03:59Speaker 02

  Are you capturing enough of the whales? And how deeply invested can a whale get? And the sort of product architectures there have created a whale dynamic in one and not in another, with a base of a subscription, which I just think is a really interesting, both in the same genre, both in some sense, the same idea, different markets, and so that's part of it. Patreon doesn't have the, at least to my knowledge, the messaging, paper messaging, pay for more engagement type features. So your product has also then set up a different set of business genetics.

• 1:04:32Patrick O'Shaughnessy?

  It raises such an interesting question around, do you understand the demand curve? And is your product architected so that there's as little friction as possible along different points of that demand curve? The way you described OnlyFans, it's like perfect. As you go up the curve, you pay more and it's low friction to do so. I think that's such a cool thing to think about.

• 1:04:51Speaker 02

  Pricing and packaging for SaaS. You spend a lot of time on it with an airtable and others that are really broad. I think really broad horizontal businesses face that question in a really interesting way, because you could use it for so many different things.

• 1:05:06Patrick O'Shaughnessy?

  It seems like that's the biggest challenge that I hear in companies we've invested in is pricing and packaging, where there's the least innate founder genius. Like in a lot of elements of a business, the founder knows who's the right customer, what should it look like, what should the product experience be. But then pricing and packages always seems to be this steel wall everyone comes up against like, God damn it, I don't know how to do this. Is there anything you'd share about navigating that decision-making process well? Because it does seem to be like the hardest thing that so many people deal with in software.

• 1:05:38Speaker 02

  I think the bad way to go about it is to hand it to, I think at any stage, but certainly in the up to a couple hundred million revenue stages I think about, is to hand it to someone in finance or someone else in the organization to go work on. I think it is a founder level, not without support, not without cross-functional work, but a founder level question that if a founder doesn't have time for it, I don't think it should be prioritized because it is deeply cross-functional and is deeply, in many senses, gene-setting. Also, there's an optimization within buckets you might get to, within a scaffold you might get to, but what's that mindset that

• 1:06:20Speaker 02

  should get created, I think is the first place to set. I think you want to think of what's the sort of global maximizing architecture, a bit of what you're saying, how can we capture a bunch of value along that demand curve? Is a question of separating out customer profiles, is a question of what's a product adoption journey, is a question of what's the full potential, is a question of what might else be on the product roadmap and potential over time. So I think it's a global maximizing framework setting exercise within which you might locally optimize. And too many, I think, start from the place of where do we locally optimize? We have this, how do we tweak it to get a little more people up selling or to monetize this segment a little better? As opposed to saying, if we look out on a longer, bigger time horizon and we ask

• 1:07:05Speaker 02

  that customer journey question or product journey alongside that, what mapping should that have? So I think those are the two starting lenses that I think a lot about with teams.

• 1:07:15Patrick O'Shaughnessy?

  One final question and then my traditional closing question for you. Do you think that there are cardinal sins in this kind of investing? Are there categorical errors that either you've made? Or obviously you've witnessed tons of investments that you didn't personally make too. Do you think that there are cardinal sins? I don't know a better term to use for it that you see happen again and again in software investing?

• 1:07:38Speaker 02

  It's sort of the things you look for and try to say yes to, but really special, founder, authentic, purposeful, intersected with a really interesting arc of product and problem to go work on intersecting with a great business, some genetics and not saying yes to those, the cardinal sinners of omission.

• 1:08:02Patrick O'Shaughnessy?

  And is price almost always the reason for that?

• 1:08:05Speaker 02

  There's a funny world where that probably got taken to an extreme in the last two years. So taken to any extreme, a rule is probably not true, certainly in the earliest stages.

• 1:08:15Patrick O'Shaughnessy?

  So I think, you know, my traditional closing question that I love to ask everyone, this is in person for those that can't see it. Rarely do I get to do these in person now, so it's more fun to ask this. What is the kindest thing that anyone's ever done for you?

• 1:08:26Speaker 02

  We're also recording it in the holiday period. Families on the mind and as we were chatting on, I have a one-year-old kid now myself. So you're imagining what it was like for your parents and you definitely get genetic payback for how much goodness or badness you might have inflicted on them. The clocks turn full circle. And I think in that you see, or at least see what unconditional lovers and the beauty of that. And I actually came to the US when I was 12. I moved over here with my mom and moved in with my stepdad. And he has given me just pure unconditional love and been a big inspiration to me and incredible person.

• 1:09:10Speaker 02

  And I think that's the most selfless thing someone could do, that level of unconditional support. And so in the holiday period and family reflection, that would be mine.

• 1:09:21Patrick O'Shaughnessy?

  How did he inspire you?

• 1:09:23Speaker 02

  He was an entrepreneur himself. He actually founded one of the very first e-commerce software companies called ATG, which ran e-commerce, I think the first version of e-commerce, everything from Delta to Best Buy, probably. And that joy of tinkering, of exploration, of building, building with a great group of people is baked into who he is. He built that company, but he also made Halloween costumes from scratch from all of us as a kid. I'm the oldest of seven, and so there were a lot of Halloween costumes to get made. He took both endless joy in that and building software that would be used by tens of millions or whatever it is of people.

• 1:10:07Speaker 02

  And I think that combo is a great inspiration.

• 1:10:11Patrick O'Shaughnessy?

  Amazing place to end. Miles, I so love our conversations. I learn a ton each time. The lesson I always learned is it depends. There's so much interesting beneath it depends. And I've loved exploring it with you again here today. So thank you so much for your time.

• 1:10:25Speaker 02

  Long time listener, first time caller. So I appreciate you indulging.

• 1:10:31Patrick O'Shaughnessy?

  If you enjoyed this episode, check out joincolossus.com. There you'll find every episode of this podcast complete with transcripts, show notes and resources to keep learning. You can also sign up for our newsletter, Colossus Weekly, where we condense episodes to the big ideas, quotations and more, as well as share the best content we find on the internet every week.