Mat Falkowski - CEO at Flock Bio

Welcome to another CEO Wisdom Podcast. We have Matt Falkowski with us today. Matt is a CEO, he's a CSO, and a serial entrepreneur with 5 successful exits, now at Flock Bio, previously at Ginkgo and such companies, founded— actually CEO of Circularis Biotechnology and a bunch of other companies. So Matt, welcome to the pod. Tell us a bit more about yourself and what you're up to.

Hey, Charles, thanks for having me. Um, yeah, so right now I'm doing a new biotech company called Flock Bio, um, and basically I'm taking sort of the 20 years experience I have of working in different biotechs in the startup space to sort of leverage DNA, RNA, and sort of AI tools and partners to sort of speed up the pace of science. And so we work with a lot of different customers and clients to sort of share our experience and accelerate sort of the work that they're doing.

Nice. And how did you start your first company? I see here you were a scientist and then you started out Invitae in 2009.

Yeah, so the first company I joined, I was working at UCSF and they had a job fair and there was one company that had a basically cardboard table with no sign other than it says Parallel Biosciences. And so, I talked to them, they explained they were a startup, they didn't explain what they do, and I said, well, this is more interesting than what I'm doing at the university right now. And so I joined my first startup kind of on a whim and, uh, really liked the sort of speed of the startups can accomplish science compared to working sort of in academia. And then after a couple years sort of working in science, um, me and a couple of my friends from previous companies decided to start Invitae. And so then grew Invitae into sort of like a $10 million company with, you know, $500 million in revenue. So it's been a really interesting journey of, you know, being from academia and small science, small startups, to even large companies.

Interesting. And in Vitae, how hard was it? I guess you went public around 2015, but you left before, right? So how hard was your first company? Because as a scientist founding a first company, sometimes it's not so obvious.

Yeah, it was actually really hard. So having had success and having a really good team, I think we had to pitch like 110 different VCs to get funding. So we were raising money in 2009 during the financial crisis and sort of raising money was extremely hard. So even though we had a good team and a good idea that panned out in terms of a successful company, it's always just, I think, really hard to raise money when people don't completely understand what you're doing or, you know, It's easier to say why it won't work than understand why it will work. So definitely a learning lesson in terms of even when you have the right idea and the right people, sometimes it's hard to share that vision with other people and have them see sort of 5 or 10 years down the road what it's going to become.

Interesting. And after that, you were a consultant, or you were— you went at IndieBio and FalconBio. How did that happen?

So FalconBio was me trying to sort of do my own thing. So I wrote a couple of grants at the R. So instead of sort of being like the big flashy venture capital path, I tried to do something more organic. So I did sort of a bunch of work and sort of like, you know, independent sort of funded stuff from like, all right, can I get this to a critical mass and make sort of funding an easier process? And then at the same time, I was also coaching teams. So having been through like 4 different startups by that point, I started to coach a lot of sort of new founders. And so IndieBio was one of sort of the early incubator spaces, sort of trying to train a new range of biotech companies. And so I worked there to sort of help share with companies what they needed to do to be successful. Met a lot of really great people who I'm still friends with, still work with. So yeah, IndieBio was a great time.

Interesting. And these founders, like what problems were they facing back then as a— and I'm pretty sure you still mentor like some biotech companies. Tech founders, what problems do they most of the time come to you with?

I think really early on it's just how to work with their team. You have a lot of sort of founder problems. You'll have a bunch of different people with different visions, sort of helping them get on the same page, understand what you need to execute on. And so for some companies, it's executing on having that first customer or having that first partner. For other companies, it's executing on raising that VC money or having that first patent. And so, you know, it can be a different problem, but I think early founders don't usually have that experience of how to focus on one problem as a team and make that progress in the next 6 months to get to that next inflection point of whether it be fundraising or customer sales or whatever.

After doing that, like, why go again and be CEO of another company, right? If you already had a couple of exits and you were just cool, like, why go back in the game again?

I think it's interesting. It's so when you work with other people who are day-to-day sort of solving problems, it makes you want to solve problems. And so when I joined Circularis as CEO, that was a company that I advised since like 2015, and they sort of got to a critical point where I saw where their technology was going, and I knew as a team they were scientifically strong, but they could really use somebody who sort of had more business experience. And so I came on as CEO because I'm like, I like what they're doing, I like the team, Um, and I see how I can then, you know, help them and how this is fun for me.

And you guys successfully sold in 2022. Is there like a pattern to exiting these companies?

Uh, a little bit. I think, you know, if you have a technology that's unique and you know who the possible acquirers are, it's sometimes a pretty clear path of saying, all right, we're going to grow this technology to a point that we're, you know, leaders in the field. but we're not too large, and that the number of people that can acquire us, you know, is let's say these 10 companies. If you start to talk to those 10 companies, a lot of them are like, yeah, this makes sense to us. Sometimes people grow their company too much. So like in Invitae's case, we got to the point where IPO was the logical path because we kind of grown past some of our acquisition partners. And then eventually, you know, you know, it made sense. Invitae eventually was acquired later on, but You know, there's different phases in a company.

And Ginkgo, how did, uh, did you come under their radar?

Um, I think I, I had talked to him a couple times at some other conferences, and then one of our VCs gave us an intro to somebody starting their biz dev department. Um, I think it's always useful to be sort of like talking to both people on the science side for collaboration, but also meeting their business or strategic people. There's always somebody at the company who has like a 3 or 5-year vision And the sooner you're on their radar, the sooner you can sort of be aligned for a possible exit or even just a partnership. Um, because, you know, it's— there's part of your team needs to focus on what's happening now in execution, and part of your team also needs to focus on where you need to be 2 years from now. And I think that's the hardest part of startup, is that you're so focused on revenue growth or funding that you sometimes lose track of where you should be 2 years from now.

Right. And they, uh, IPO'd when you were there, right?

They IPO'd just before, and they were a SPAC. And, um, that was kind of a problem because, you know, at the time most SPACs were coming out, they were getting hammered. And so their stock would start high and crash pretty bad. And so when we did acquisition, I was a little bit worried about that. So we ended up doing sort of partial stock, partial sort of other deal. I think that was an important thing to recognize is sometimes the company acquiring you, you have to be careful of, you know, are you being acquired by a company that's all stock and might go down during your lockup period, or are you being acquired by a company that's growing and that you want stock and that you'd rather not have cash. Like, if Nvidia buys me tomorrow, like, yeah, I'd be happy to take stock. On the other hand, you know, if you see some company out there, stock has been shrinking year over year, you always have to worry about that acquisition.

And they laid off a bunch of, uh, their team, right, at some point?

Yeah, so within like a year they laid off— they did a large layoff and they did a large layoff again. And so They had the funding, they sort of had a vision. Um, I think there was a lot of problems in terms of the execution, um, on the executive team of seeing like what some of the market segments really needed. Um, and so you just had a lot of middle management, a lot of properties, and a lot of equipment, and they didn't have sort of that clear execution vision on like, you know, profits or, you know, things like that.

Interesting. What, uh, what led to Flock? And you again, getting back in the arena, you can't stop it.

So after my team got laid off, they asked me to sort of join them and start a new company, basically doing what we did, but on a much smaller scale. And so I think when you're a large company trying to service everybody, you end up with a lot of sort of bloat and sort of this expansion of like costs that you can't reduce. But when you're a small team, like my team of 4 people, you can reduce costs and you can be highly effective. And so they asked me to join and we decided to focus on a very narrow sort of aspect of what our group was doing. which is building DNA libraries. And we did a really good job. We have a lot of customers. Our revenue growth has been great, but, you know, really done good science. So solve a tractable problem and the customers will start showing up. And that's what we've accomplished.

And Flock is like, from what I understood speaking to your co-founder the other day, right? It's more of a service, obviously have like data, right? But it's a team, right? It's sort of consulting if you want, Is there as much scalability to that one as there would be to your last 5 exits?

Yeah, there is scalability. And so it's when you build out the tools and the team to do things, then you just find other sectors where you can accomplish that. So like, we have other projects that we'll probably eventually look into where we internally develop assets. And so at this point, that's not the primary goal, but once you start understanding the field and what can be solved with the tools that exist, we can find a partner. And in a partnership, we can create, you know, value that way. So instead of taking all the risk of raising a large amount of VC money and setting a goal to be somewhere in 3 years, we can work for 3 years and in the meantime notice what's needed and then find the right partner who can help us execute on that. Because if you do it all yourself, you have to buy a lot, you have to build a lot, you spend a lot of money. But we're working in a system right now where it's so easy to work with a lot of different companies. Like, as long as you can manage that project and trust those partners, You don't need to build everything yourself. There's not a lot of scenarios where you need everything to be in your own ownership.

And what changed in biotech since AI happened?

So since the AI happened, I feel like what the scientists can execute directly has improved. And so in the past, you would have sort of your scientists, your computational biology team, your programming team. You'd have a lot of different teams. But what AI has really enabled is my team members whether it's sort of handling the front side of the customer or final reports or doing programming, they can use AI to sort of supplement their knowledge and skills without having to go through another person. So that means things are getting done quicker, things are getting done without having that extra layer of somebody interpreting someone else's work. So if my person can take their work directly into deliverable to a customer or solve that problem by writing code, that saves sort of that telephone game with people making mistakes in communication.

Interesting. What's been challenging about that specific company?

I think one of the challenges has been convincing our customers to just sort of trust us a little bit more. I think people come in with ideas from papers they've read or what somebody else told them to do, and, you know, it's sometimes hard to convince these scientists or CEOs that, you know, we've been doing this for 10 years and that they— the reason why we do things one way or another is because we've seen things fail. Like, you can save $1,000 here, And at the end lose $50,000 when you do the, you know, the next phase. And so we're always trying to sort of educate people on ways to save money long term or how to like plan for higher success rates. And it's been frustrating sort of having people, well, we'd rather do this. And you're like, you can do that. And because we're a service, we'll do what they ask. But I wish sometimes that they would like approach it the other way of, well, if you were doing this, how would you do it? And then say, okay, I see why. You do it that way.

And as a guy that's been into sales as well, how hard is it to sell to these people? Is it a long sales cycle? What, what is your close rate? Give me some stats there.

Sure. So our sales cycle originally we thought would be closer to sort of like 4 weeks or like, um, in the shorter side. It's probably been closer to like 2 to 4 months. And I think it's just people have an idea and then before they have the budget to execute on it or the time to execute on it, I think there's a lot more In-between steps. And when we can, we help our customers. We've had customers who've come in and within 2 weeks we've got them on path to build the library they need because they trusted our input enough to say, okay, I'll listen to you and let's get this done sooner rather than to do 4 months of planning. Like, data now is better than data 5 months from now. And that's a balance that, you know, you have to make on an executive level.

Interesting. And how do you get these folks through your pipeline? What's your main lead gen method?

Definitely there's some word of mouth. We've gone to a few small conferences like SynBio Beta, and then LinkedIn has been very effective of reaching out to people both we know and sort of random people, companies that are more recently funded. So obviously things like Crunchbase tell you when somebody's been funded, and then reaching out to those companies, they're more likely to have budget or growth versus companies that have done layoffs are usually not worth reaching out. Because they're not spending.

Right. What about outreach, you know, like on LinkedIn, for example? Have you done some of that cold email? What's, what's like been the feedback on that level?

Feedback's been generally good. We've tried to keep things really short. Like sometimes I'll just talk to people about the space. Sometimes we'll agree that it's just not a match. I don't believe in a technical field like mine that you need hard sales, so it's not like push, push, push. It's more like ask if this is a match and then wait for when they need that. Occasionally maybe a reminder, but—

Is it CSOs usually the target ICP?

Not necessarily. I think on the smaller companies it can be CEO or CSO, but sort of on the mid companies sometimes it's just a senior scientist. somebody who's looking for a solution and they recognize that it can save both time and money. Um, so it, it can be both. We've— we assumed in the beginning it would always be high level, but we found that, um, some of these smaller companies, they give their lower employees more flexibility to sort of find solutions. And so you'll find people who just want to show that they're, you know, good employee and that they're finding solutions for their company to succeed.

Interesting. And the companies that use this data, can you tell us like the specific use case and what they would do instead of using your solution before?

Yeah, so a lot of people would do it themselves, and so something that we might be able to do in 2 or 3 weeks, they might kind of fiddle with for 4 or 5 months sort of trying to do it themselves. Other times people would simplify it, and so there are some vendors out there who will send them just very basic versions of what we do. And so instead of doing the ideal experiment, they'll simplify their experiment to something that another provider might provide to them. But once again, it would take a longer time for them to get it, and they're usually sort of not getting the same quality of data because they cut a corner for simplicity reasons.

Interesting. Hmm. So saving time, right? Like, how do you pitch the product specifically? Because saving time doesn't make a lot of sense to everyone. Some people just don't value their time. Do you put like a numerical figure to the potential time that they would save and an ROI? And how do you base that on science and facts, you know?

Yeah, so I have a A slide in my deck that I always try to point to, because when I tell somebody that a project might cost $40,000, they're like, well, we can do it for ourselves for like $35,000. And I'm like, you can, but if your team spends 100 hours doing this project, that's 100 hours of lost productivity if they can be working on another project. And so what they need to understand is if that 100 hours is $30,000 of labor plus the $35,000 to do it, that project internally costs them $65,000, not $30,000. And I think a lot of people who haven't been sort of on the accounting or CEO side, like they don't understand sort of the budgets of their own team. And so they don't realize that work not done by your employees needs to be calculated into that cost. That's a lost cost. Like you're losing money when your employees are doing stuff that someone else can do for you for 10% more. Most biotechs, like even a small company, like you're burning $100,000 a month on employees only. And so when you start to try to save $1,000 here, $1,000 there, you don't get that, like, the burn rate of your employees alone usually exceeds the savings you can get on a lot of these things.

In large companies, right, like life science companies, they wouldn't need that data, or still, are they still a target fit for that?

They are target fit. They need the data. Finding the right sort of champion at those companies is always the challenge. And so they have sometimes more clear roadmaps, people's roles are more defined, and so finding the person with the right role who has the right authority and will actually go to their boss and say, we should do this, that's the sort of challenge. When you find them, they can be great. You can have large projects, recurring projects, but it's sometimes a challenge to find that person at a large company who will go sort of outside their comfort zone and say, we should be doing this instead of, you know, doing it in-house.

Interesting. Um, and I would guess, uh, yeah, this data, is it like regulated in some way? Like, how do you make sure that it's compliant and, uh, it follows their, their expectation?

Yeah, so a lot of the data that we work with, um, it's not patient-related, so it's more of just like keeping the data private or confidential. Um, So when I was working at Invitae, for example, we had to do a lot of HIPAA compliance and sort of like protecting, you know, patient data. Most of the stuff we're doing now is a little bit more research-oriented, and so it's more about protecting that data for the customer so that they can generate IP or, you know, you know, work on it. They don't want to make sure the data doesn't end anywhere public. And so there has been a lot more concerns in terms of both AI and how when you share data and you store everything like Amazon, everything else, like People are starting to worry a lot more that their data is getting sort of like, you know, used by AI and things in ways that they don't want to. And so that's kind of been an interesting avenue to understand of like how to protect your data, not just in the sense of like privacy, but, you know, when you're using all these tools and all these partners, how can you make sure that they're also protecting your data and also that some of your data sources aren't showing up somewhere they shouldn't be?

And how far are we away from AI scientists that are fully autonomous and just running their own experiments?

From the sort of like design perspective, a lot of people are already using the AI tools, cloud science and stuff like that, to design experiments and do things. From the standpoint of actually physically doing the experiments, there's definitely workflows that companies have automated. They're called cloud labs. And so they'll basically build a lab where you can put in a protocol and then it'll run that protocol. There's a lot of limitations on that. So what a scientist can do is much more flexible in terms of using different tools or reagents or setups than these cloud labs are designed to do. I think these cloud labs will continue to expand, and in the next sort of like 5 years, you'll be able to do a large percent of your experiments on cloud labs. But if you're cutting edge, you're always sort of changing the variables, and a lot of those variables are not as flexible with some of the robotic systems. So I think we You know, are much further away from that. I think the humanoid robots, so they start to, you know, get into that gray area where like if it can physically behave like a human and work in a lab where things are not designed to be robotic or automated, like that starts to change that option of it can do things that are more custom.

And what's their limitation? Uh, can you program creativity and, I don't know, gut feeling into an AI?

I don't think you have to necessarily pre-program those things. I think it's a question of like how you can teach it to, you know, think outside the box or intentionally fail. I think sometimes, you know, people complain about AI always trying to make you happy and not say no. They think there's a sort of necessity in biology sometimes to intentionally do experiments where you expect it to fail. So instead of always designing your work to succeed, design some percent to succeed, some percent to fail, and you'll sometimes learn more from those sort of negative controls than you will from the positive controls. But that's even a problem with young researchers. I'll see people coming into the lab who have a habit of doing experiments for things to succeed, and instead of taking half of that work and saying, well, half of my work I know is going to fail, but it's going to be educational in how it fails— and that's a thing that we learn over time as well.

What's next for you and the company this year?

Continue to sort of do growth. The number of customers we have has been really good, and so we're sort of trying just to expand out our lab and be able to handle more customers. Continue to talk to new partners in the space. To what you said, you know, I always see sort of that further vision, and if we find the right partners, I can see how our company can grow to significantly beyond just sort of this custom services company into something that's, you know, creating sort of novel content products, IP partnerships, really pushing the envelope of what can be done with sort of combined AI tool DNA.

I love it. Where can people find out more about your company, Matt?

FlockBio, the entire website. We're posting usually on LinkedIn a lot of stuff. Yeah, that's probably, you know, give me an email. I love to talk science to people, whether they think they can use us or maybe use us. I always enjoy Both advising companies and sort of whether they're partners or they're just people