How do you revolutionize venture capital in a rapidly evolving technological landscape? For Chris Farmer, founder and CEO of SignalFire, the answer lies in using AI to uncover exceptional talent and drive groundbreaking investments.
In this conversation, Chris shares his journey from private equity and entrepreneurship to building SignalFire, a $2.1 billion AI-native venture fund. His unique approach centers on using data and technology to navigate early-stage investing, where the focus isn’t solely on ideas but on identifying exceptional founders who can adapt and pivot. SignalFire’s proprietary AI platform, Beacon, analyzes over 650 million people and thousands of companies to predict where the next groundbreaking talent and opportunities will emerge.
At the earliest stages of a company, Chris explains, the idea is almost always wrong. The real challenge is backing heat-seeking founders—entrepreneurs with the agility and vision to navigate dynamic markets. This approach has enabled investments in companies like Grammarly and Grow Therapy, which are using AI to disrupt massive industries such as healthcare and legal tech.
Chris sees AI as offering one of the greatest commercial opportunities in history. Unlike traditional software that improves productivity, AI-driven solutions can directly replace labor, cutting costs at an unprecedented scale. Take Grow Therapy, for instance: the platform streamlines scheduling, billing, and marketing for independent therapists, empowering them to focus on clients while scaling their businesses efficiently.
This focus on talent and AI-driven insights is reimagining the future of venture capital. Chris highlights that smarter, more efficient systems are critical to supporting founders and scaling transformative ideas. By combining high-conviction investing with AI-powered data, SignalFire is reshaping how venture firms discover opportunities and build for the future.
Transcription:
Steve Brotman: Back to Driving Alpha. I’m Steve Brotman, and today we’re joined by Chris Farmer, a true innovator in the venture capital world. Let me give you a quick rundown of Chris’s background. He’s the founder and CEO of SignalFire, which is a cutting-edge VC firm that’s redefining investment strategies with its data-driven approach.
In 2013, Chris honed his skills at some of the most prestigious firms in the industry, including General Catalyst. He earned his bachelor’s in international relations and business at Tufts, and at some point, he was actually on Wall Street, co-owning a company and their private equity group before founding an investment advisory service for angel investors and private equity funds.
So his journey through the VC world actually started at Bessemer Venture Partners and then went on to General Catalyst, where he led seed stage investments in game-changers like Coinbase, Discord, and Stripe. But Chris isn’t just about big firms. He got his operational experience too, managing a firm called SkyBitz as their VP of management and business development.
With over 15 years in the industry, Chris has developed a unique blend of entrepreneurial experience and investment acumen. Today, we’ll dive into how he’s leveraging his expertise at SignalFire to revolutionize the VC landscape. Hey, welcome Chris.
Chris Farmer: Thank you so much for having me. It’s a pleasure.
Steve Brotman: Super, I appreciate your coming on. Let’s get started a little bit. So, your path, your backstory, and how you got into venture capital—how did that winding road take you to SignalFire? Just briefly, if you could share that.
Chris Farmer: I come from a relatively untraditional background. I grew up overseas and was very interested in international relations. So I went to Tufts and the Fletcher School of Law and Diplomacy, which is, of course, where every good VC goes.
From there, I was very interested in the implications of the Internet on labor markets globally and international trade. So I wrote my thesis in college on that, discovered venture capital in the course of that journey, and never looked back. I had never met a venture capitalist at the time, but it was the perfect intersection of my interests in entrepreneurship, investing, and strategy, and all those types of things.
Steve Brotman: Right? And I think we first crossed paths when you were at Bessemer, if I’m correct.
Chris Farmer: Yeah, it’s been quite a journey. I was at Bessemer, oh, whatever—’06 to ’08, something in that timeframe, maybe ’05 to ’08. It was a fantastic place to learn the ropes of venture capital. The alumni network there is spectacular, and there are lots of great people there as well.
Steve Brotman: No, it’s remarkable. I would say that you’ve always struck me as someone who’s a very deep thinker, not just about investing in a particular company but also in many ways about the venture capital industry overall.
As we get older, I think you start thinking about that a lot more. Tell me, how did the idea for SignalFire coalesce?
Chris Farmer: So, when I was at Bessemer, I had run a company previously that was using data in the sort of industrial supply chain world as a SaaS platform. When I came to Bessemer, I was covering wireless because it was one of the most international sectors when the iPhone and the App Store came out.
I thought it was sort of a lost cause to try and eyeball the App Store. So I created an algorithm to create a leaderboard, effectively, on that. It worked out really well in the beginning for gaming apps and social media, etc., but it was pretty clear that I would need to dive much deeper and cut the data with lots more data sets, which I wasn’t going to be able to do at a more traditional firm.
That sort of launched me on my journey of pulling on the data thread and seeing where that took me in reinventing a new approach to the venture industry.
Steve Brotman: So heavy on data analytics, right? And machine learning.
Chris Farmer: Yeah. Honestly, in my early years, I was in Boston during a period where there were almost no consumer hits. I was scouring the universe trying to find great new companies, but it’s very anxiety-provoking when you’re not seeing all the best companies. You’re not in the right rooms.
I had frankly run the wrong company, gone to the wrong school, and was in the wrong geography. I didn’t really have any competitive advantage other than market mapping and trying to think of where the world was going. So the application of data was a way of me getting leverage, making sure that I was seeing the best things so I had a fighting chance.
Candidly, coming from a point of relative weakness at the time, because I hadn’t gone to Stanford or been part of the PayPal Mafia, Google, or whatever was the “academy companies” where all my peers were starting companies, I had run a company in DC. I was in Wellesley Hills at the time in a suburb of Boston, working for Bessemer.
It was very stressful to try to think that you were going to be in all the right rooms at all the right times. I used data as a crutch just to make sure that we were tracking everything and seeing the best opportunities.
That started a journey. When I left Bessemer, we went and studied over 170 different funds, interviewed over 500 founders, and deconstructed the models of everyone from quants to corporate accelerators, the best-of-breed investors in venture and private equity. We really went deep—down to the tech systems they used, the source of founders, how they promoted, who they hired, all these types of things.
It was one of the most detailed studies of venture ever done as far as I’m aware. And from there, we sort of came to the theses that led to the foundation for SignalFire.
Steve Brotman: Interesting. I mean, breaking out—just leaving your firm and doing all that work. I mean, did you have a team to help you do that? Or was it just you? I’d love to understand more of the genesis there.
Chris Farmer: Yeah. I mean, I had six researchers for a year, but I did the poor man’s version of that. I found people who were back at B-school. Folks like Eugene Chung, who actually was the first hire at Andreessen Horowitz for their platform, was my lead researcher, and he and I collaborated later. He ended up running our platform for a number of years.
And just sort of people who wanted to get into venture. 2009 was a period where there were no jobs, so you really had to go the extra mile to stand out. I just went deeper and deeper down the rabbit hole. I had plenty of free labor—people who were interested in venture—and I used it to its fullest.
Steve Brotman:
I’ve kind of gone through a similar journey with Alpha—just sort of a unique model. What was the reception from LPs? Did they just immediately say, “Boom, we’re in,” or was that a difficult process?
Chris Farmer:
So, the first chapter—a lot of people… I literally— I won’t say the name—but I met with one of the founding partners that I thought would be the most forward-leaning from one of the marquee venture firms. The meeting lasted literally seven minutes. He was so offended by the idea and the concept of what we were doing, despite the fact that he’d founded a quant wealth management startup and was at a major brand-name venture firm.
But it was offensive—some of the concepts that we had that have since proven out. Definitely got our share of naysayers, and most people thought it was insane to try to apply data to seed at the time.
I ended up partnering with General Catalyst—Hemat Taneja, David Fialco, and Joel Cutler—who had been very involved in launching companies like Livongo, Demandware, and Kayak in-house. They had a very talent-driven strategy to find the best talent as the atomic unit of what we were tracking, which ultimately starts this catalytic process of starting companies.
I approached them and said, “Hey, give me five million bucks. You don’t have a seed program. I’ll help you launch into seed, I’ll help you find EIRs, and we’ll build companies together.” Then I was going to find somebody on the West Coast to do that. They wanted to be my national partner, but I wasn’t willing to do that without somebody senior from the firm moving out to California.
So that’s how it all played out. I moved to California, joined as a venture partner, and they were my sole limited partner in the first two funds. We brought together a motley crew to join me. Eventually, as now-partner at GC, but he would never have interviewed there. He came in through my seed vehicle.
We worked super hard, we were scrappy, and we got very lucky. We seeded five trends: Stripe, Coinbase, Zapier, Zenefits, and lots of other companies—Segment, etc. It was quite a run.
They’ve done an amazing job of building GC into a global franchise, but at the time, it had just been in Cambridge prior to that. They supported me in launching SignalFire in 2013 when I decided to spin out.
Steve Brotman: So they really funded your pilot, right?
Chris Farmer: Yeah, no, it was a great win-win. They were my sole limited partner in the first two seed funds—the pilot funds, effectively, as you said. They then did a tremendous job of investing upstream in companies that came out of that seed ecosystem we developed.
In addition to the Stripes that were huge wins for the firm—where they’ve now invested, I think, something like $700 million—other things like Snapchat and Airbnb came out of that presence in California and those seed activities. Even though we didn’t get into those deals until a little bit later, it worked out super well all around.
It allowed me to prove concept. So then, when I went to go raise our first institutional fund in 2015, I was super fortunate to have some fantastic LPs like Horsley Bridge that anchored our first fund. They’ve been just fantastic partners ever since.
Steve Brotman: Well, they’re based in Boston too, right?
Chris Farmer: They were originally spun out of US Richmond. They definitely have a presence here in San Francisco—I’m not sure about Boston—but they sort of have folks around the country.
They were very quantitative in the way they looked at the venture world and are sort of the premier fund of funds and one of the most thoughtful LPs in the industry. I was very fortunate that we aligned on our vision of a data-driven approach to early-stage investing.
Steve Brotman: What would be the pushback? Having more data seems like an obvious advantage—like, more data.
Chris Farmer: Today, yes, but I would say that people felt that data would probably be more helpful at the late stage. At seed, how much data is there? This is a talent-driven business, a vision-driven business, and at the very earliest stages, there’s just not enough data to pick up any pattern recognition.
That was the biggest pushback. And candidly, nobody wants to back a data-driven firm before you’ve built the platform. But how do you build the platform if you don’t have capital to do so?
We had to do a lot of unnatural things. We raised money for the firm and the operating company itself. We are structured like a tech company—a startup, in that sense. A lot of people felt like we were going to be a Mattermark or a Bloomberg, not a fund.
We advised hedge funds, corporates, and everything to amortize the cost of the infrastructure. We spend north of $10 million annually on data engineering efforts—not including comp and carry. It’s very expensive to do this. Our Amazon Web Services bill was half our management fee. The credit card data set was more than the entire management fee.
It’s a hard effort to get off the ground. A lot of people were skeptical. If we put everything into data, how would we hire great talent on the investment side? So it was a delicate balance.
Steve Brotman: You know, Bob Olis, I don’t know if you knew Bob at Sigma?
Chris Farmer: It’s been a long time, but yes.
Steve Brotman: Right? I was involved in some deal—RecycleBank—that ultimately didn’t do so well. I actually didn’t invest, which was a good thing. But we had a big debate about that internally: is it the founder or the idea?
Bob felt that it’s really easy to change the management team, but it’s not so easy to change the concept or the opportunity set you’re going after. Would you agree with that or disagree?
Chris Farmer: We have a very strong bias to not changing out the management team. I think you lose a lot. In my experience, until something is relatively later stage and commercialized with post-product market fit, you tend to get very linear execution—very good execution—if you hire correctly, but not necessarily the bobbing, weaving, and pivoting to continuously find markets that are evolving that you see with entrepreneurial founder DNA.
Chris Farmer: And so I’m a big believer in backing founders in interesting markets. They need to be heat-seeking missiles, figuring out how the business models are evolving. The AI landscape, for example, is transforming under the feet of all the companies building on top of it. You’ve got to be super agile.
General ideas lead to follow-on products and movements into adjacencies. You need to be agile until you’ve really built something that has broad appeal and scalability. That sort of discovery journey requires very specific DNA—it’s that risk-taking, fast iteration, those types of things. Executives tend to be better at operating and linear execution, while founders excel at the fast-twitch agility needed in the early stages.
Steve Brotman: I didn’t mean to suggest that changing out management teams is a strategy or a good idea. It’s more about relativity, like in your weights and data models. If you have an A-team but a B or C concept, you can’t make a C concept work with an A-team.
You can have the most pivoting, enthusiastic founder in the world, but if the concept itself isn’t right, it’s very hard to turn that around. But if you have an A-concept and a B or C team, it’s easier to upgrade the team and help them execute. That’s really more of what I was thinking about.
Chris Farmer: Yeah, totally fair. We have a heavy bias toward backing the best heat-seeking founders. At the pre-seed and seed stages, we assume the idea is always wrong—99% of the time. That doesn’t mean it’s completely wrong; it just means you have to tack or pivot a little bit to get there.
You don’t have to transplant completely to a new market. We’re looking for exceptional founders in interesting markets where disruptions in technology have the potential to bring impactful change.
There’s usually an evolution: you’re dealing with corporate customers, who will give you feedback on the pain point. You may be close to the right pain point but not the exact one. Or you may have a wedge with a medium pain point, but you use that entry to solve for a bigger pain point.
As a result, we tend to layer in capital. We back the absolute best talent we can find in markets we think will be most disrupted and pour fuel on the fire as these companies find product-market fit and go-to-market motions that work.
For us, it always starts with the absolute best talent, and, in some cases, great data sets—particularly with AI—that are differentiated and allow you to build better products.
Steve Brotman: So your data analysis isn’t so much on the end. Maybe that’s the catch. You’re evaluating founder capability—and that’s where the typical VC would hang up the phone and say, “You can’t do that.” There aren’t enough data points to evaluate that. Are you looking at the kid’s 8th-grade transcripts from their teachers? Or maybe you are—I don’t know.
Chris Farmer: Let me answer without going into too much specificity. It starts super people-centric. We evaluate many, many parameters of people’s backgrounds—yes, going all the way back to their academic careers, anything they’ve done on the job, open-source projects, patents, and all sorts of things.
We’re looking at learning velocity. We start with great talent in interesting markets. Over time, the way we use data evolves. It’s not the same datasets or parameters at every stage of a company’s journey.
At the earliest stage, the focus is extremely talent-oriented. Over time, it starts to shift toward customer behavior patterns and commercial execution elements—some of which are team-oriented, some of which are more go-to-market or execution-oriented.
The weighting shifts as you get to different stages of a company’s development. Even then, it depends on the sector. The weighting of your AI team is completely different for a fashion marketplace than for an AI infrastructure company or an LLM. You have to have the right type of talent for the specific use case.
We break markets into thousands of different subcomponents that are different for each of them. You need massive-scale AI and ML infrastructure to do these types of things because they can’t be human-tuned.
Steve Brotman: Interesting. You’ve had a lot of successful companies. I mentioned a few of them earlier. Can you give me one or two case examples just to bring it all to life? It doesn’t have to be very specific—like my baby and exited company, Getaround, is one of them. Not the best example because we were investors there too. But maybe you can tell me what went wrong there someday soon.
It would be great to connect this a little bit more.
Chris Farmer: I would say it starts much more with discovery. At the pre-seed or pre-launch stage, the number of companies you’re looking at is obviously orders of magnitude bigger. We’re actually looking at pre-formation entities. It’s something like 80 million entities—some of which become companies, many of which don’t.
It could be an open-source project, an interesting person leaving a job, or a founder still at a company who hasn’t yet left to start their next venture. Statistically, based on our algorithms, we project which people are most likely to start companies and succeed as serial entrepreneurs.
The first step is finding the needles in the haystack and getting the timing right on when to reach out. We’re tracking hundreds of millions of people and tens of millions of entities.
As companies incorporate, raise capital, and go through the phases of formation, the number of entities we’re tracking starts to winnow down. By the time you get to Series A, B, and beyond, there are fewer companies to track. That’s when you go much deeper into the operating parameters of the company.
The discovery problem becomes less challenging. The differentiated view into a company becomes more important. There’s more data to analyze—like customer quality and retention, reviews, press and PR signals, or platforms like G2.
We can see customer credit card transactions, purchase frequency, and retention rates. We track API usage, open-source activity, and team quality—like whether they’re recruiting and retaining great people.
All of this becomes a way to track a company’s heartbeat. The moment they’re building momentum—even under the surface—is the kind of signal we pay close attention to so we can get ahead of the market.
Steve Brotman: So you’re pre-seeding and seeding a lot of companies, and then as they progress, you’ll select the ones you want to really double, triple, and quadruple down on based on the data and their performance. Is that accurate?
Chris Farmer: We select, and the market selects. The ones that don’t find traction typically don’t raise Series A, B, or large rounds of capital. The ones that graduate successfully attract more upstream investment opportunities simply because they’re raising more capital on the back of customer success and sustainable metrics.
There’s an inherent Darwinian process to all startups, regardless of the firm’s strategy. But we have the capability to go the full lifecycle with companies. While seed is our general point of entry, we’re not a seed fund. We have the ability to stay with companies all the way through their journey, and today, we have as much as $100 million invested in a given company.
Steve Brotman: What’s the most surprising thing that you’ve discovered—something everyone thinks is true, but it isn’t?
Chris Farmer: I think the most surprising thing has been how much window dressing there is when it comes to data and how little depth many firms have actually gone into. Being natively born as a data-driven, AI-native firm is radically different.
The analogy I use is with quant hedge funds: all the best quant hedge funds in the world started as quant hedge funds. Nobody started as a traditional firm and transformed themselves.
Now it’s become almost vanguard for every firm to say, “Of course, we’re using data.” But that usually means they’re buying data from a third-party source. That’s no different than Fidelity buying a Bloomberg terminal. Yes, Fidelity uses data, but that doesn’t make them a high-frequency trading firm.
The most surprising thing for me is that while many firms—Google Ventures, Social Capital, and others—have tried this, most of them have stalled or shut down their engineering efforts. I expected there to be a stronger competitive set by now, given how much consensus there is that data matters. But competition has actually diminished, which is the opposite of what I would have predicted.
Steve Brotman: Interesting. Have valuations become a consideration? Like, you could have all the green lights in your model—great data, great company, everything looks good—but the valuation is just insane. Do your models factor that in, like when something is priced at 100x revenue?
Chris Farmer: Yeah, that’s been the most frustrating part. Our models crush the venture industry in terms of returns theoretically, but translating that into reality is constrained by the ability to get deals done at reasonable terms.
Between 2018 and 2021, when the markets went crazy, we actually lowered our cost basis in companies by over 20% while the rest of the market was paying two to three times more than they had previously.
As valuations inflated, we shifted earlier—we went more aggressively into pre-seed, seed, and post-seed rounds, and we stopped doing Series B and later-stage deals. We effectively traded valuation risk for company risk because we felt valuations were outpacing the probable outcomes for these companies.
We stayed disciplined. If the market is overpaying, you need to get ahead of it so others pay up for your investments later. If the market cools, you can then buy more advanced companies for the same dollar.
Having that perspective on market cycles is critical. You have to avoid getting caught up in consensus thinking. At the end of the day, it’s about staying disciplined and achieving risk-adjusted returns.
Steve Brotman: Wow. Have you found that you’ve been able to repeal or at least bend the power law curve?
Chris Farmer: I can’t get into too much detail, but I’ll say that our loss ratios relative to the rest of the venture industry are dramatically lower. Our capital concentration in our best companies is dramatically higher.
The power law dynamic does shift as a result of that. We’re able to keep our losses lower and put more resources behind the winners, which, hopefully, will yield the better returns we’re targeting.
Steve Brotman: Right. And the outside folks you used to sell your data to—were they mostly in public markets?
Chris Farmer: We stopped doing that. In the early years, we had to do it to reach critical mass. Seed funds don’t generate enough cash flow from management fees to make heavy investments in infrastructure.
We structured ourselves like a tech startup in those early days, but it became too complicated to explain to LPs and navigate conflicts of interest. Today, we’re focused entirely on being a fund.
That early experience showed us why it’s so hard for traditional firms to shift into a data-driven model. It’s like GM trying to become an autonomous driving company. All the moats and incentives are in the wrong places. You have to reinvent your DNA, your compensation models, and your entire approach.
The firms that succeed in this new AI-driven paradigm will need to be born from the ground up, like SignalFire.
Steve Brotman: Do you use your models to choose sectors as well, or is that more of a bespoke process?
Chris Farmer: It’s hard to predict the future perfectly. Our models help us react to sectors where companies are gaining traction. We can see clustering patterns and movement.
The models are less predictive at the earliest stages, which is why we like to get in at pre-seed and seed. But they do show us where talent is going. That’s one of the signals we focus on.
We also do a lot of macro work ourselves to understand opportunities and the nature of technology. One of the biggest secrets to our success is that we’re not just technology investors—we’re technology product builders.
We’ve built full AI product suites for our portfolio and for our team. We’ve been using AI and LLMs since the earliest days, and that gives us a bare-metal perspective on where the technology is working and where it isn’t.
Steve Brotman: Chris, this has been fascinating. I feel like we could chat forever. Thanks so much for joining me today.
Chris Farmer: Thank you so much for having me. This was a lot of fun.
Steve Brotman: Chris Farmer, CEO and founder of SignalFire—thank you for being on the podcast.