Episode 127: Jim Ferry of Volition Capital on
Moats, Defensibility and Investing in the Age of AI

Episode Transcript   

Shiv Narayanan (00:09)

All right, Jim, welcome to the show. How's it going? Yeah, excited to have you on. Obviously Volition is a good partner of ours and we've had other folks from the firm on the podcast, but why don't we start with your background and a little bit more about Volition and then let's go from there.

Jim Ferry (00:11)

Thank you, thanks for having me.

Jim Ferry (00:24)

Yeah, for sure. So Volition is a series A, series B growth equity fund. Think about us investing in businesses with five to 50 million of revenue or so. Our typical check size is in the 15 to 60 million range, I'd say. And we're investing in our fifth fund, 675 million. We're tech generalists, you know, as a firm and each partner here tends to kind of have their own areas and sectors of expertise. So I joined Volition 12 or 13 years ago at this point. And a lot of my time has been focused on kind of high volume transactional type businesses. So that can be across sectors such as advertising technology, payments, FinTech, digital assurance, supply chain and logistics, marketplaces, whether B2B or B2C.

Shiv Narayanan (01:20)

And in those markets, a lot of these companies are dealing with large total addressable markets and AI is having a big impact as well in those types of companies. So I'm just curious, like, what are you seeing inside these businesses and where are some of the core value creation areas that you're focused on?

Jim Ferry (01:38)

Yeah, specifically with AI. Yeah, I mean, it's interesting because I think we're at this point in time where AI is so new and fast evolving. There's almost a question of our historical portfolio that we've invested in over the last 10 plus years. And a lot of them were, those businesses were founded in a non AI world. So they're having to adapt and we're spending a lot of time, you know, trying to create some institutional knowledge across the portfolio. So for instance, we have, you know, frequent webinars where our portfolio companies are just showcasing some of the cool stuff that they've built internally to make their processes more efficient or how they're using AI to kind of code and ship products faster and trying to share that knowledge base as well as getting the portfolio together for our upcoming leadership summit that we do once a year to get all the management teams together. Obviously the topic this year is AI and a lot of it is kind of real hands-on workshops of here's what people have done, here's how you can do that at your own business. And then there's kind of this new flavor of companies that we're seeing now that have been founded over the last year or so and they tend to be from the start more native AI businesses. And what that means is they're using AI as a way to solve problems in an agentic way without just hiring more people, which I think was the traditional way. So a lot of them are taking the approach of, historically the traditional sales model was you have this many people with a specific quota, 75% are gonna hit the quota potentially, and here's what we're gonna pay them. That's the math that's gonna make the business work down to the bottom line. Now we're seeing a lot of different strategies. Like people are using AI agents to post on their social media pages and they're becoming similar to an influencer or content creator within their specific vertical that they play in. And then they're kind of every once in a while posting about the product that they have that's relevant to their audience. So that's kind of a new age way, I think, for distribution that we're going to see more.

Shiv Narayanan (03:50)

Are you focused with these companies on evolving more of their go-to-market in this environment or on overhauling product or rethinking product in an AI-first world?

Jim Ferry (04:02)

I think it's all of the above and ad hoc to what the company needs. I do not have an engineering or coding background, so I am less helpful and hands-on on the actual product development side, which is why we try to create some institutional knowledge across our portfolio, as I mentioned. But I would say if you think about the tech leaders that we have, they tend to be the early adopters of AI. So it's not like we need to push anything onto them from AI adoption. It's just happening organically in the tech side of these organizations. I do think that the sales and marketing customer success organizations do need a little bit of a push in some companies because they're not kind of wired to solve problems via technology. Historically, they've just said, hey, we got to hire another customer success person because we onboard another 10 customers, for example. So I think in those scenarios outside of coding and shipping product where we're trying to add more value and really my job is about pattern recognition, what I'm seeing in not only our portfolio, but the thousands of companies we talk to and meet with on an annual basis.

Shiv Narayanan (05:13)

Are you seeing on the go-to-market side a decline in volumes? The reason I ask this is we're seeing this as a common problem inside of a lot of companies that historically have relied on channels like paid search or SEO and organic traffic. And now as queries and search volumes are moving to AI platforms like ChatGPT, and Gemini, or even zero-click searches, there is lesser volume and lesser inbound traffic and conversions and pipeline coming to these companies. So I'm curious, especially with the types of businesses that you're investing in that are broader based, like are you seeing that as a trend as well?

Jim Ferry (05:49)

Candidly, I can't say we've seen a lot of performance degradation within our existing portfolio, you know, based on, hey, we used to be SEO driven and now it's all LLM, GenAI search driven. However, I think a lot of these companies weren't relying on like SEO inbound marketing. They're less of them were kind of a PLG product-led growth type company that might rely on that. But what I am seeing is a lot of these companies, because the budget for testing AI tools is pretty large these days. You know, I think given the pricing models are a little bit different than historically, they tend to be more usage based. People are very willing to test and try new products. So there's almost a virality that's happening somewhat organic within a lot of these native AI businesses that we're seeing, which allows them to scale at unprecedented levels with a small number of employees where historically in a pre AI world, when I think about Volition sourcing, we have a traditional growth equity sourcing model of kind of reaching out to thousands of founders and having conversations and seeing if there's a good fit, that's how we get our deal flow. Typically you kind of needed 25 or 30 employees to get to like a five million run rate, which is when we tend to get involved. You need a certain amount of engineers, a certain amount of salespeople, a certain amount of execs. That's just not the case anymore. I think we just saw our first two person billion dollar exit that happened recently. So that's no longer a signal because of the organic virality that a lot of these products have because people are talking, hey, what AI tools are we using that you found really efficient, et cetera.

Shiv Narayanan (07:38)

So what is the new signal? Like how do you figure out if it's the right time to invest when these companies are smaller or fewer resources? Like how do you figure that out?

Jim Ferry (07:49)

I think it's how you define smaller. Their revenue could be larger. It's smaller in terms of infrastructure and the actual team. So then it's kind of a scalability question. It's a great question. Especially on the consumer side, for instance, we've seen companies that go zero to 20 million run rate. And a lot of these people, as I mentioned, are testers of the product. And you haven't really hit a renewal cycle yet. And then there's somewhat of a mass churn event because it was a nice to have, not a need to have. So from my perspective, the ultimate question in every investment committee right now is around durability. What is the defensibility and durability of the product in the long run? And there's a lot of ways that you can achieve that. One would be first party data. Another one could be distribution, could be some type of integration that you have, that's a non-public API that gives you an advantage. Could be a system of record network effects, et cetera. So that's the ultimate question that we're asking on every company. And, you know, when we do get to the point of, okay, this company went from zero to five million run rate in three months, historically, we would have had two, three years to run the analysis. You're taking a little bit of a leap, but there are ways to diligence that, such as customer reference calls to understand what are the usage patterns, what are the login patterns? And how likely are these people to renew based on is this a nice to have or a need to have and transformational to their day-to-day workflows.

Shiv Narayanan (09:21)

Yeah, I completely agree with this durability and defensibility take because it's kind of hard to figure out if a company will stand the test of time without having some of those elements. Can you expand on those? You mentioned first party data distribution, having some type of proprietary integration system or record. And then also this idea, because everybody's talking about SaaS companies being disrupted and agents and all those things like. How are you future-proofing your existing companies to make sure that they do have something defensible?

Jim Ferry (09:53)

Yeah, it's a great question. Kind of a two-part question. So within the existing portfolio, I think that's where, as I mentioned, it's less of a push down from us to say, you need to adopt AI. I think everyone is cognizant of that, every company needs to become an AI company and reinvent themselves if they were founded in a non-AI world. And we've had a lot of founders that are making that transition really well. I think some are just moving at a velocity and pace that are faster than others. But for the most part, everyone is aware of this, it's here. I'll give you a couple of examples. Like there's one company in my portfolio that has been really strong at adopting AI and they kind of looked at all their third party — it's a pretty sizable business — but they looked at all their third party software vendors that they use. And realized that there were a few that they could probably recreate internally on their own and maintain them. And they cut a million dollars of software expense on an annual basis, just by kind of recreating the software. Kind of an interesting thing that's happening is after that, I asked all my other portfolio companies, hey, do an audit of all the software vendors that you use. And are there any that you can just recreate internally? And what we found is it's more of the higher price ticket items that people focus on. There's, you know, somewhere might be more SMB type pricing. That's five to $7,000 a year. And it could be easily replicable, but it's just not worth the headache because our companies aren't going to recreate every single software solution that they use. So I'm finding there's a convenience factor. Even internally at Volition, we're setting up a lot of our own, you know, shared skills and we have OpenClaw bots running from a sourcing perspective and have a bunch of kind of shared artifacts and so forth. But if someone came with something out of the box that's better than what we're piecing together, given we're not technologists, we'd probably pay for it if it was purposefully built for growth equity. So I think convenience is something that people are potentially overlooking when it comes to these new AI software companies.

Shiv Narayanan (12:14)

Yeah, I think all of those are super important. I think especially having a moat around the system or the data that cannot be really duplicated or kind of vibe coded away as an advantage is a huge one. What about on the people side? Like, how are you guys looking at organizations and potential transformations in light of all of this? Like, are you working with CEOs to adjust their teams or think about headcount or margins around that differently now?

Jim Ferry (12:45)

Yeah, for sure. So we have a talent team internally here at Volition that helps hire world-class talent for our portfolio companies, but also helps them kind of think through what you just mentioned of team structure and org chart. And our companies are high growth businesses. So it's not like they tend to not have a lot of bloat at this point. I don't think it's like none of our companies are really doing mass layoffs, but they're slowing down on hiring and saying, hey, we may not need to hire anyone new to grow this in 2026, maybe even 2027 where historically part of the budgeting process was to, you know, it's like 20 new hires or whatever the math works out to. So it's more about keeping the employee base flat as opposed to laying people off these days. Where we're seeing the biggest gains is on the R&D line. I think that not only are they the early adopters of AI just because they've kind of, they're technologists in a way. But I think that that's where pretty much every company is seeing the most leverage today from the AI coding tools. It's starting to trickle down into other line items like customer success and so forth. But I think everyone is just becoming more efficient, which means you know, CRO or the entire C-level executive suite may be able to do more than they were doing historically because they freed up a little bit of time. And that doesn't mean they're sitting on their hands. They're just taking on more responsibility.

Shiv Narayanan (14:23)

Yeah, I mean, all of that resonates. How do you vet for these types of things? Because one of the interesting things about your firm is that you're looking at earlier stage companies and founders. So how are you vetting for that as founders are pitching you and like thinking about all of these different elements, whether it's the org side, the product side, or how you find efficiencies or how do you scale go to market or the implications of AI? Like how are you vetting founders at the investment stage?

Jim Ferry (14:53)

Yeah, there's no silver bullet on it, but there's specific questions and you can kind of get a feel for how the founder thinks about utilizing AI over time. And it's little things like in their financial projections that they might show you, you can track the headcount growth over time. And what is their distribution strategy today? How are they utilizing AI in their day to day to make themselves and their team more efficient? Is this siloed across the organization where a couple people are doing cool things or have they set process-wide organization-wide skills and rules that everyone can use on the day-to-day basis? Ultimately AI is just moving so fast that I think everyone has anxiety that they're behind no matter how evolved you are. If you miss one day on Twitter to get all the updates of what's going on, you feel like you missed a month or six months. So we're trying to find people that are tinkerers in AI. And our general philosophy at Volition from an investment standpoint is we can't invest in AI unless we understand AI and know how to use it. So we are encouraging everyone here to just play around, not only for business use cases, but on Mondays we have Volition AI Labs, we call it, where everyone in the entire firm gets together and we show some practical things, from a demo perspective, that make our lives easier, but also just some fun aspirational things that we've built just to learn. Cause I think that's it. Like you need to play around to learn and then it starts to unlock the possibilities and capabilities that you can do to change your day-to-day workflows and make yourself more efficient.

Shiv Narayanan (16:37)

And so what can founders do to kind of resonate with you guys in that way or show that they are doing some of these things to either future proof their companies or actually uncovering what those growth avenues would be for their businesses?

Jim Ferry (16:52)

Yeah, I think number one is the most obvious — just building agentic workflows into their actual products, like that has become table stakes. So that's number one. Number two is show me don't tell me. I'm a visual learner, but I want to see a demo of what you've built that has made you more efficient because I do think that there's a lot of people that are all talk no game when it comes to AI and a lot of them talk about, yeah, we're using Claude for this and that, but it's really around email automation and nothing that's kind of groundbreaking. But I want to see someone who has multiple OpenClaw agents running autonomously on their behalf to accomplish specific tasks. We have portfolio companies today that have automated the entire BDR process. They're using agents to scrape prospective customers, use a bunch of third party and first party data to reach out to them, send them something that's extremely custom based on their social media page, so if they like skiing you might put a meme about skiing in their email outreach — all that is automated. So their sales team shows up and they have calls booked, instead of having to go prospect them themselves. Little things like that add up and give us comfort that someone is going to be forward thinking about it.

Shiv Narayanan (18:24)

Right. So I love that example. I guess my question on that would be that that's something that theoretically could change in six months or nine months or not be as effective. And it's not really defensible. Right. So is that enough for you guys as you're looking at a company or is the AI or agentic approach, does that need to be built into what the company is actually selling on the product side?

Jim Ferry (18:55)

So you have to know, you're right that it might be commodity in nine months from now. And that's what I mean by, not everyone is doing that today. So if they are doing that, it gives me comfort that this person is going to be up to date on all the capabilities of AI, be able to evolve with it over time. That's what I'm looking for. It's not someone who builds something and is static. It's someone that can evolve with how fast everything's moving with AI. And we kind of joke here internally at Volition, but it's half true that there's five things that matter as a product, market, management, management, management. So we spend a lot of time with founders to feel like, you know, regardless of the business that they're running, that you'd want to back this person at anything that they did and that they're relentless.

Shiv Narayanan (19:42)

Yeah, I think that's a phenomenal insight — it's a signal on the capabilities of that management team and how forward thinking they would be. So I'm curious, like what are some other areas that you're looking for those kinds of signals? Like the sales process side. What are some other — or maybe if you have some examples of places where it's really jumped out that this company is ahead of the curve or on top of the latest developments.

Jim Ferry (20:07)

Yeah. You know, I've had founders move on from a tech leader that might be 30 years in the industry, super talented coder, but is just not changing — changing someone's habits and status quo can be difficult. So for some people that aren't evolving with AI, they're moving on from people that they don't think are forward thinking on AI. And obviously you want to keep all of these great employees, but it's the reality of the situation that you might fall behind if you're not tinkering and playing around with AI. And it is fun, it's almost like a video game to play around with it on the weekends as you're watching TV or something like that. So that's one. And going back to like hiring, I think it's people that are asking themselves the question, do we really need to hire someone or can we solve this problem using AI? That's another one that gives me comfort when people are thinking that way. Sometimes you might need to hire someone. I'm not saying AI can do everything, but taking a step back to even ask yourself that question is really important.

Shiv Narayanan (21:18)

Yeah, I think that's great. Are there any metrics or accountability measures or ways you're using to measure whether a company is actually leveraging these tools or is AI forward? As an example, you should see some form of margin expansion if you're deploying this in the right way. I'm curious what you think about on the metric side.

Jim Ferry (21:39)

So it's interesting, I actually don't know if you're gonna see margin expansion right now. And the reason is because what we're looking for is token usage. So, you know, I think Jensen and NVIDIA said like, you know, if you have a $250,000 engineer and they're not spending $250,000 in tokens, then they're not forward thinking from an AI perspective or working hard. You know, maybe that's an extreme example, but that's a great indicator of how much are they actually spending on AI. And because the companies that we invest in tend to be a little bit earlier stage and they're growing fast, they're probably just shipping more products, moving faster. It doesn't necessarily mean we're experiencing any margin expansion on the bottom line yet. Because you might be offsetting some of the human capital costs with usage of AI and token costs, but at the same time, you're moving twice as fast from a product perspective, if that makes sense.

Shiv Narayanan (22:47)

It totally does. I guess the token usage at some point should feed back in terms of incremental revenue or additional speed. So how do you look at that as a time horizon in terms of returns?

Jim Ferry (23:00)

It's interesting. I don't know if we have the right metric for that yet. The way that traditional SaaS metrics are defined and here are the best practices, I think we're still in the early innings of trying to understand the ROI and that stuff. Even historically, it's amazing when I go to board meetings and we'll have 50 slides on the sales and marketing efficiency and metrics and R&D might have a couple slides and here's what we spent and here's what we shipped. It's been a little fuzzy and hard to really put an ROI on the products that the R&D team is creating and what the work that they're doing. Is it really cleaning up technical debt or building new products? So it's always been a little fuzzy. So I candidly can't say we have the best metric to measure the ROI in that, but ultimately what I'm looking for is that the products that we are creating are either accelerating the sales funnel or making sales easier or upsell potentially.

Shiv Narayanan (24:16)

Yeah, I think there is some sort of a downstream — or there should be across the board. If for example, you're using it for sales efficiency or marketing efficiency, as you're spending more long term, you should see improved cost of acquisition or more pipeline and deals. But I guess it's a little tricky too, because you have to spend money upfront on the tokens or on it — it's almost like a J curve of learning and onboarding the organization into that. So how are you encouraging companies to take on that cost because they have to be willing to almost slow down to move faster or take on the cost of actually taking on some of this work?

Jim Ferry (24:57)

Yeah, it's funny — I've said that in a few board meetings recently where we need to slow down to speed up because it takes intentionality to change the status quo of here's how we used to do things, here's how we can do them going forwards. So I completely agree that in the long run, you should experience efficiencies, but if you're a five million run rate business growing to 10 to whatever the math is, you're probably too small to start to see that because you're just, you know, even with AI, you do need people. You're hiring for growth and probably spending a lot on the token costs and usage based pricing. However, when you get to 50, 100 million, you should start to, depending on your growth rates, start to see some efficiencies as you continue to scale. And ultimately, the market pays for growth. So when I think about our portfolio companies, if you're kind of growing under 30%, you better be profitable because that's what the market's going to look for. If you're growing 50, 70, 100% plus, it's okay to be burning because it probably makes sense to, and the market's going to value you on a revenue basis.

Shiv Narayanan (26:07)

Yeah, I guess as you're talking about all this, where my mind goes is that if everyone starts to do this in every market, then is some of this like a hype cycle where the amount of growth every company needs to achieve to make up all of this is like, it's not even possible in particular verticals and industries. I'm curious, like, how you think about that, because it feels like some of the growth — I mean, there's growth to be had, but more companies are fighting for the same market now because it's easier to enter into spaces or code a competitor and kind of show up in that space.

Jim Ferry (26:42)

I think there is a lot of hype right now. Not every AI business, just because it's an AI business, deserves a ridiculous multiple. We've seen companies go zero to 20 million run rate in seven to nine months and get wacky valuations. And someone's underwriting that that growth is going to continue over the next three to five years. Even if you experience that growth, it's not a slam dunk that you're going to continue at that rate. Law of large numbers, numbers get bigger, it's harder to kind of grow at the same percentage base. So I think that there's a lot of, we're almost coming back down to a normalized zone. Early on in this cycle, maybe a couple of years ago, anything that was AI based was just getting a ridiculous valuation. I'm talking to folks on the West coast who were saying seed deals are getting done at 200 pre, which is ridiculous. And now I feel like because every new business that's being founded is now an AI business, they're kind of coming back down to, well, if every business is an AI business, we can't pay a ridiculous multiple for all of them. So they're back to like a normalized zone. What I think is getting really hurt right now are traditional SaaS companies that have all been bucketed into this category that they're screwed because AI is going to overtake that category. And I don't necessarily agree with that. I think that there's some categories, obviously, that are going to be impacted more than others. But if you're a system of record or it goes back to some of the moats that I mentioned around first party data, et cetera, I think that those are probably undervalued and are well positioned to adopt AI as long as they're kind of forward thinking relative to new entrants coming into the space.

Shiv Narayanan (28:27)

How are you underwriting your own companies or prospective investments to make sure you're not inflating the possibility there, right? Because companies might be growing faster or there's an opportunity in front of them. So I'm curious, what's your approach to underwriting there to make sure you're not overpaying or over indexing on a particular asset?

Jim Ferry (28:47)

It's a great question. We talk about it a lot, especially if you look at kind of public SaaS multiples right now, they are very compressed. And as I mentioned, I don't think that is a reality. I think there's going to be a good amount of money made in a lot of the public software companies whose fundamentals are really strong right now. And they're all pivoting to add AI to their products. And a lot of them are going to be fine. Some of them are going to be displaced over time. So it's almost a fine in the happy median of, there was a period where software multiples were trading like crazy. We're in a compressed zone and there tends to be — given we're in the private market space — a six to nine month lag in the private market where we play relative to the public markets. So it's very interesting where I've had deals over the past three months where we've been on them and sometimes I think it's a business that's going to trade on EBITDA based on the sector that they operate in and the financial profile and growth. And someone pays a wacky revenue multiple. And then the flip side, there's times when I think it should be the inverse and no one really shows up to bid on the company. So we are in a very weird market right now. From Volition's standpoint, we've never been the type of firm that's going to chase the hypest, hottest startup that has the chance, you know, six months in, you feel like you overpaid. We're a little bit more patient capital and looking for strong businesses with kind of low loss rate risk that have the potential to be kind of impact deals. And so we take a concentrated approach to the fund. So one thing we've been doing a little bit is on some of these AI businesses that are a little unproven — you can write a smaller check and then try to get more capital in over time in the form of a capital call option potentially, or maybe a super pro rata right or something. So you're not putting a huge check into a business that needs to prove out some of the infrastructure despite the scale that might be ahead of the actual employee base and infrastructure that they've built.

Shiv Narayanan (31:08)

Yeah, I think that's a great takeaway. I'm just curious, like, are you moving potentially slower or at least being more selective with the companies? Because it's not just about finding a good company, it's about finding it at a fair price or a good value. So I'm curious how you look at that.

Jim Ferry (31:23)

Yeah. I mean, we're seeing really good businesses now that I think are trading at, you know, native businesses that I think are trading at like realistic multiples, not 40, 50x, which I'm just not going to pay. That doesn't make sense to me. So I think, like I said, people are being grounded. There's also this other bucket of company that we haven't really talked about where Volition has made a lot of investments over time and these would be companies that I put in the category of less disrupted by vibe coding and AI. Because ultimately code is no longer defensible because it's easy to spin something up. So we talked about some of the moats that you can have, but there are other types of businesses. Like I've invested in a couple of marketplace businesses. The technology for those has never really been the moat. It's been balancing supply and demand and creating a virality effect for growth. We have a lot of hardware-enabled software businesses in our funds. It's like a software that's tied to a hardware product. And I don't care how good you are at vibe coding, you cannot vibe code a piece of hardware. So that has become the moat. And there's a lot of businesses in our fund, historically, where those models that a lot of the big PE funds that we tend to sell businesses to were somewhat out of favor. And now those are the type of companies they're asking about because they recognise that there's a built-in moat relative to AI. So it's kind of funny how things change like that.

Shiv Narayanan (33:05)

Yeah, yeah, I think all of that is — I think a lot of firms that are thinking about investing and trying to pick companies and setting up their portfolio companies for success, there's so much to take away from that. Jim, we're coming up on time here, but before we close off, if people want to get in touch with you or Volition, what's the best way they can do that?

Jim Ferry (33:24)

Yeah, my email is [email protected]. If you're a founder and you want to chat, I am happy to have a call and see if there's an opportunity to work together. I'm also trying to be a little bit more active on Twitter so you can follow me at @jimferryvc.

Shiv Narayanan (33:43)

Awesome, we'll be sure to include that and all the links in the show notes. And with that said, Jim, thanks for coming on and sharing your wisdom. I appreciated the fast paced nature of the conversation and also a lot of the AI insights. I think a lot of firms are trying to figure this out and I think your take on a lot of these topics was refreshing. So I appreciate you doing this.

Jim Ferry (33:59)

Thank you for having me, I appreciate it.