Sudarshan Sridharan (Fion)
The CEO of Fion Technologies discusses talking to customers, doing hard things, and the future of natural disaster prevention
Welcome to the inaugural issue of Master Plan, a series of conversations with deep tech founders doing really hard things.
Sudarshan Sridharan is the CEO of Fion Technologies and also co-founded Gen Z Mafia. Before that, he started a VR company, a nonprofit, and was even featured on CNN Money at age 16. Suds was kind enough to give us a detailed look at his journey building Fion without a heavy technical background, some advice on moving fast, and his thoughts on doing hard things.
Sudarshan introduces Fion
At Fion, we try to prevent fires before they start. Once they start, we use satellite imagery to detect them faster than anybody else, we predict where they’re going to spread, and then we estimate how much damage the fires are going to cost.
We’ve got 3-4 data scientists, a sales guy, a sales intern that works under him, our CTO that manages the entire development, and then me.
Early customer discovery
When we were early on in the product development cycle, we talked to a bunch of potential customers. Initially, it started off as fire departments, then graduated to forestry services, then state-level people like sheriff’s offices and organizations that were coordinating multi-county regional fire response efforts.
And then we finally started talking to insurance companies and financial institutions and trying to figure out what they want in the product. Once we’ve figured that out, we worked on figuring out what we should triage and prioritize, and what was the key product offering that we needed to provide in order to have an MVP that we can sell.
Once we'd figured out what we needed in the product, we started building it and we started trying to do sales before we had something. So a lot of our days were just calling people, talking to people, understanding why we weren’t being able to sell.
The day-to-day of building a deep tech company
A lot of my time also goes toward answering emails. It seems trivial, but we’re looking at big partnerships and most of our sales stuff gets done over email and 15-minute Zoom calls. So it’s a lot of emails, used to be a lot of hiring, used to be a lot of product stuff, now it’s more about trying to get sales done.
Fundraising also looks very similar, where I have meetings every day. Today, I started off with meetings because I do angel investing on the side, and then I spend a lot of time trying to understand how other people are running their companies and what I could be doing better.
I do standups with my entire team throughout the week. I just ask them what help do you need and it’s my job to make that help appear. Let’s say we need access to an API, or maybe a data source goes down, or something needs to be faster. I’m the one going in and finding the right engineer or the right partner that can solve that problem.
I’ve been limiting myself to about 3-4 meetings per day. Meetings are only for critical things, although I occasionally have office hours 12 hours a week where people can sign up to get help.
Other than meetings, what you don’t see on the calendar is basic clerk work. We’re not a massive company, so it doesn’t make sense for us to hire a general counsel to fill out the paperwork for us.
On trusting your technical team to set achievable goals
I let my technical team set their own deadlines. I tell them “Hey, we need something, you guys get it done.” If we have a hard deadline, I’m like, “Hey we need it done by this. What can you get done?” and help them spec it. If we can’t get something done by the deadline, we triage and prioritize to figure out the most important things we need for whatever the project is.
Beyond that, I’ll say “here’s a hard deadline, get it done, how long is it going to take to get this done?” If we don’t have a deadline, let’s cut our estimate in half and get the first version out in half the time.
We are very deep tech, so a lot of this you can’t say “hey let’s ship half the features.” Things usually take 3x longer to ship. My job comes from figuring out where the redundancies are and where my team is still thinking that they’re in academia instead of business. I try to help them adjust to that. They understand we need to move fast and they understand we have a lot of things to do, so the team themselves are going to set good achievable goals.
How to move fast in regulated markets
Fion sells to governments and insurance companies, which normally have lots of bureaucracy involved that prevent people from making decisions quickly. How do you move fast in an environment like this?
A lot of it comes down to your mentality. The traditional sales cycle is 6-9 months for governments. For insurance companies, it’s probably longer. Where we really shine is we’re not just sending the intro email and calling it a day. We’re looking at it and asking questions like hey, this is a new avenue/industry/segment. How can we attack it? Who do we know? Who should we be talking to? Can we get 10,000 emails out? Is this a numbers game? What are the different segments?
A long sales cycle doesn’t stop you from iterating fast. For example, our API is our core product and from there we have a dashboard and other structures because you don’t just sell a bare API. You need to structure and package it for a customer.
And so for the API, we’re asking ourselves: what is the fastest way to get that out the door? What information do our customers need? Our customers simply need to know where fires are and where they’ll go. That doesn’t require a dashboard. It can be an email or text notification at first, then you can build a dashboard backward from there.
It’s about how fast you can iterate on the product, how fast you can talk to potential customers and figure out what they need and don’t need. You don’t just need to be physically faster, but you also need a mindset shift of constant iteration and forward momentum.
Leveraging conversations as a learning shortcut
You’ve talked about how before building Fion, you’ve been doing insane amounts of research on fires. What’s your learning process like?
I talk to people, I used to read all the time but ever since I've found Twitter I’ve decided to stop reading textbooks. I could have learned about fires by reading research papers, and I did. I read so many research papers, and I typically never do that.
But most of my learning came from calling up people who’ll never give us a contract in the first place, like a fire chief. Even though we couldn’t sell to them, they still spent 2 hours just educating me about every part of what they do, all their problems, and the general workflow.
I used to get on 5-10 calls per day. I used to not close a single one and it was really embarrassing because I wasn’t sure what I was doing wrong. But part of that is because we didn’t have a product, and we didn’t know how to sell without a product.
How to sell without a product
Mockups and visuals. You can even simply ask customers about their problems, then tell them: what if we gave you a product that just solves those problems? Would you use it?
If yes, cool. Let’s sign an LOI, let’s make your life easier.
It’s pretty simple. Tell me your problems and I’ll tell you how we’re gonna help you solve it.
How to look at fundraising for a deep tech company
Fundraising isn’t about valuation, it’s about looking long term and asking how does the cap table look at series B and C, what are we building for, and what is that money going towards. Right now, our investors are investing in the team, our knowledge moat, and the idea of what Fion could be. They’re not investing based on sales numbers and metrics.
And also we’re a little bit different—especially with me being non-technical. Some people might think we’re a pure SaaS company, but we’re actually deep tech. So concerning valuation, people need to be coming at it from the viewpoint of a deep tech company. We’ve been in development for 10 months. Even if Google or Amazon wanted to throw their PhDs at it, they’re still going to run into the same problems we did.
Having a bunch more smart people in a room doesn’t change anything when the technology is evolving on a literally daily basis. Machine learning is literally improving every day, in terms of where the boundaries are and what’s possible. Our team is just as fast and agile as any other big company’s team will be, and we’ve got a first-mover advantage.
What the first version of Fion looked like
Version zero looked like a consumer tool that notified people if a fire was coming close to their house. Our angel was Josh Browder from DoNotPay and he built a very successful business out of $3 MRR. So I was like, that would be a great place to start and prove out the efficacy of our model.
What ended up happening was that it wasn’t worth diverting all of our resources into a consumer product just for the sake of revenue. Even if it gets us from zero to one, it wasn’t going to be sustainable enough to take us from 1 to 100. So we decided that it was a complete waste of time and we went back to the drawing board and started talking to fire departments to understand what they actually want.
Learnings from previous companies
Before building Fion, Suds founded Second Reality Interactive in 2018—a company that built a virtual hangout space for watching Twitch in VR with your friends. The vision was to create a space where anyone, regardless of location or physical ability, to gather with friends or have conferences and meetings. Second Reality shut down in March 2020.
I still don’t know 90% of the things I think I should know now, but I knew zero of the things I should have known running Second Reality, which was really rough. I didn’t have a good understanding of how you get a product from being an idea to being a product.
I started developing that understanding over time with other ancillary projects I’ve worked on (like SaveMAPS), but it wasn’t until Gen Z Mafia where I was doing a drop every 3 days and also PMing every single one. In my head, I’ve basically got it down to a systematic process of how to get from idea to spec to building a team to launching, and then from launch to generating hype. Those things I was able to codify in my head even if I couldn’t explain it.
And from the business side, I mean, coming in completely cold totally just knocked me off my feet because I had no idea what I was doing with Second Reality.
At Fion, we’re an 8-figure company and we still didn’t have some basic things done, so we’re still learning. But Second Reality was more of a way to learn how do you build a product, how do you get it shipped, what does that mean, how do you fundraise, and why is that so hard? And then from that first failure comes a lot of learning experiences.
Thoughts on VR and the tech industry as a whole
Right now, the tech for VR isn’t there. A VR headset is running on a chip that’s at minimum 2 years old. You won’t get an ARM processor until 2022 (if they’re even thinking about that). The emphasis hasn’t been on making the tech exponentially better, but on making it incrementally cheaper or more accessible. And so those are the real problems in the VR space, and I don’t see the economies of scale part getting there anytime soon.
All the PhDs are at Google wasting their lives, and anyone else that can build anything is working on ads for Zuckerberg or wasting their life at Apple. The people that could be building better hardware to move an industry forward are not doing it because they’ve been bought out.
How to make a trillion dollars
One of my biggest theses is “if you wanna make a trillion dollars, you’re not doing that if you’re making an app that lets you push a button to get something delivered at a ten-billion dollar valuation. That’s fundamentally kneecapping yourself.
For me, I always want to be swinging for the fences because you’ve only got to be right once. And when you’re right once, you’re making a 100 billion, or a trillion dollars on the high end—not just a billion. I don’t really see myself doing something besides Fion in the future because there’s a lot of maneuverability.
At one point, we could expand into computing services. We also need to throw up our own satellites to get our own data faster than what NASA can give us, and there’s a lot you can do with a satellite that greatly expands our product offerings and ancillary services.
What makes a good founding team for deep tech
The thing about deep tech is that most of what people would consider “deep tech” isn’t some new groundbreaking thing. It’s more about taking things that have been in academia for X amount of time and getting that from a technology readiness level (TRL) in the 4-6 range to a TRL of 9. And then taking the business side from zero to ten.
This is where you need an operator—or someone who understands how to do sales—to work with academics and bring people together around a vision and goal rather than just saying “We’re gonna pay you $200k for this”.
The biggest mental shift you need to make is that academics are super risk-averse. In academia, your research is your reputation—it’s your name. When you’re in tech and you need to move fast, those different mindsets clash. And so it’s more about sitting academics down and trying to figure out the happy medium where they’re comfortable slapping their name on it but also lets you keep moving at a startup-level pace.
How to get more people to do hard things
I think it’s about incentive alignment and risk. I think most people don’t have the conviction to go all-in on themselves. Something I’ve realized is that people don’t all want to be amazing. They just want to do their thing and die. 95% of people are just here to die. It’s about finding the right people who want to take the swing and then figuring out how to give them the resources that empower them to try and swing for the fences. And that’s basically what a VC does.
I think it’s important to look for people who are self-starters. You can probably develop or train for it, but I don’t think finding people who have potential and trying to change their entire mentality and mindset works every time. It’s a lot easier to take someone who’s hungry and try to scale them up to a certain level rather than take someone who’s smart and has no ambition.
Fion’s Secret Master Plan
I think our trajectory would be to build all the software and get the physical infrastructure (like satellites) in place to verticalize fire response and understand everything about fires.
Once we’re there, why stop at selling to insurance companies and hedge funds? Why not become an insurance company? Why not start trading derivatives and futures and start looking beyond fires?
After that, we want to become a data provider for extreme weather. Why not look beyond fires for other natural disasters and extreme weather events and become a data provider, like Clearbit for natural disasters?
You can follow Sudarshan on Twitter at @itzSuds if you haven’t already. If you want to be notified every time we send out a new interview (we have a few more coming up soon), you can subscribe here. For any questions, feedback, or requests, send them to me via Twitter.