In the following exchange, Sudowrite’s co-founder, Amit Gupta (@amit), moves admiringly across a divisive binary of starting up, revisits some promises and struggles, and documents the inner wisdom (and the still open questions) he has gathered over 20 years of building.
— How the past has informed Sudowrite’s unique path
— The origins of Sudowrite
— Saying no to institutional investors and burnout
— Contending with a downturn and a galloping software category
— Making time for the adjacent possible
— Running lean and the forces that dictate that choice
— Sudowrite’s (surprisingly) first few GTM steps
It’s interesting how much stigma is attached to how we talk about some of these outcomes. When we say there should be a 3rd option — aside from 1) building a hypergrowth startup and 2) building a lifestyle business — there’s an implicit hierarchy.
I believe you can get to great outcomes in many different ways.
You just have to pick the one that works for you.
Here’s the background for our particular choices with Sudowrite.
My co-founder, James, started a company called Parse, back in 2011. He sold that to Facebook in 2013 and then spent several years there post the acquisition. Eventually leaving to begin writing science fiction.
I’ve had a similar path.
I founded a company when I was 19. In college. Dropped out. Raised some capital. Went through the dot-com bubble (and the burst). Had to cut the company down by 50%. Twice. It was a grueling journey back to breaking even.
But we made it work.
Then I went back and finished school. When I got out, I had already experienced building a company in the “take a lot of money and try to grow very, very fast’’ mode. So, I did a couple of (deliberately different) things next.
I helped someone start a non-profit. Which, obviously, wasn’t a venture-backed business.
Then I founded Photojojo, which was a bootstrapped business. It just grew with me. I only hired people once we had the money in the bank to hire them. It was on a very different growth trajectory. And I sold that in 2012.
The reasons for selling are complicated.
I actually bootstrapped, because I never intended to sell. Photojojo was something that I really enjoyed building. I wanted to keep doing it. And why not? It was profitable within the first few months and was doubling or tripling each year.
Then in 2011, I got a diagnosis of Leukemia.
That kind of shifted the whole dynamic. It was no longer a question of doing this forever. Because forever might be just a couple of more years. So I decided to sell and try other things.
I, too, ended up writing science fiction.
That’s how I met James. We were in the same writing group.
We were writing together. Critiquing each other’s work. For a year or so. And then GPT-3 came out. A huge language model that changed everything for us.
Because it showed us, for the first time, that there was a possibility to create something special for writers, quite like what Photoshop had been for photographers. They had to make stuff in the dark room, remember? Taking hours just to develop the film.
The advent of digital cameras made photos instant. And with Photoshop, you could do anything you imagined with photos. That brought the power of computation to the art.
Writers, on the other hand, have only had spell-check and grammar tools.
Not much else.
When James and I began playing with GPT-3, it was really a toy for us. We weren’t interested in starting a company. We just thought we’ll build something cool for our writer’s group and play around a bit. And some tinkering ensued.
Then we showed it to a few other writers we respected and their reactions were very strong. They were blown away by the things this could do.
Even from that point, it took a year before we decided to turn that tool into a company. We just kept building for ourselves and our friends. We weren’t charging for it. We just weren’t ready to jump into anything.
Until it became more and more clear that there were a lot of things we wanted to build for the tool and that we were going to have a hard time doing that without more people.
That’s when we settled on raising a seed round. This was late last year.
And we were incredibly intentional going into that decision.
James has a family. 2 little kids. I live in Hawaii. I surf, I want to keep writing. We’re both used to this almost post-scarcity life. The most scarce thing for me, and to be honest, for anyone, is time. And we’re devoted to people we love. We didn’t want to take away from that by having all our time sucked into a new business.
So we set out thinking, we’ll do this.
We’ll raise. Only if we can raise from people who are aligned with our mission to make this a different kind of company, one that’s not just hopping from one funding round to the next, attempting to be a Decacorn.
Only if they’re okay with that.
Which meant that we said no to institutional investors because it would have been hard for us to resist fundraising again. Instead, we raised from almost 100 individuals. The round came together really quickly.
The Sudowrite team is still pretty small. It’s just 5 of us. All remote.
No one is in the same city as anyone else.
I have this to-do that recurs every so often. With a goal we had set for ourselves in the early days, which was to get to a point where we’re working no more than 20hrs a week. We’re not there yet. It feels like I’m working more every week. Not getting closer to that point either.
Still, it feels fun. That was the case with Photojojo as well. I often treat my work and life as intermixed. I don’t like having boundaries between the two. But I like having balance.
That’s the challenge for us. How do we make sure we’re not burning out? Because I definitely burned out before. I think I’m getting better at paying attention to the signs. My girlfriend, who is also a founder, is also very good at spotting them.
She’s both supportive of me putting in the time when I have to and also great at asking, “are you sure you want to do this?” To which, the answer is usually, “yeah, you’re right, I’m pushing a little bit too hard.”
That’s really helpful.
Some of that awareness has also come with age and maturity. It doesn’t feel like the end of the world if something doesn’t get done this instant. You know there’s always more than you could ever possibly do, no matter how big your team is or how many resources you have.
It’s not a question of doing more. Everyone says that. Nothing new there. But it’s still hard to make that decision to not do something and be okay with it.
That’s a growth area for me.
Giving up more. Delegating better. Learning to do less.
When we raised, the funding climate was pretty frothy. Then earlier this year, the market really fell off a cliff. Making it much harder to raise. Especially later rounds. Probably even seed rounds.
Then, in the past month or so, generative AI blew up almost overnight. We’ve been working in the space the whole time but now, suddenly, these developments have captured the zeitgeist.
And because we’re one of the few companies that have been at this for a while. There’s just a lot of attention. We get VC emails every day. And we’re pretty much just saying no right now.
But we’re open to exploring.
If we can make something happen where we get more firepower, more cash in the bank, and don’t have to give up control, don’t have to give up a board seat, this might be worth pursuing.
Our growth is strong and our burn is low. We have 3-4 years of runway, even if growth stalls completely. In other words, things are going well and we don’t need money.
So, why raise?
There are two reasons we’re considering it: 1) We started our company (Human++) with the intent to create an array of tools to help people be more creative with AI. Sudowrite is the first. More money would enable us to experiment with the next ones sooner.
And, 2) no one knows how long the economic downturn will last or what the funding climate might look like in 2-3 years.
Considering our unique choices in the current funding environment, are there opportunities that we’re giving up?
We had that conversation with a bunch of our investors when we were passing through San Francisco last month. One of those chats was with Ev Williams, who started the incubator that became Twitter among other things.
He’s no stranger to trying a lot of different things at once and seeing what sticks.
We spoke with him about the current state of Sudowrite. That we’ve been recording consistent month-over-month/week-over-week growth and have a clear trajectory.
But we have all these ideas for other things we could be building. They came up all the time. Based on the things we’re scoping for Sudowrite. Extensions of technologies that we’re developing there or broader opportunities we’re noticing.
Some of these are well in the writing space, some are completely outside of that. So we wanted to learn, how should we be approaching this as a lean, early-stage team?
Does it make sense to take time from this project that’s working and growing and work on this thing that doesn’t exist yet?
Ev was very encouraging.
He said that all the good things in his life had come from following his instincts and just playing with and trying things out.
We’re at a point where the company is beginning to feel a little bit more stable. Like it’s really working. And we should pour more resources into growing it.
At the same time, it seems important to set aside time, every couple of weeks, or each month, to work on something completely different. Just to see where that experimentation takes us.
It might result in a new product that’s really cool, or something that flows back to Sudowrite and improves the core product as well.
Ev had a great metaphor for this.
He talked about how when you open a door (explore a new idea) you don’t just see what’s in that room, you’re going to see other doors. Sometimes it’s not the first door or the second door, it’s a door deep inside the building that’s the one you really want to get to.
If you never open the first one, you’re never going to get to THE door.
It really feels like that right now with the new models. The more people play with them, the more possibilities they’re finding. They’re discovering stuff we didn’t know about 2 weeks ago or even 2 hours ago in some cases.
It’s a very exciting time to be in this space and discovering and opening doors.
The first year of Sudowrite was actually an uncomfortable one for me.
Mostly because my first two companies got to product-market fit very, very quickly. And looking back I strongly feel that, in both cases, the first versions were the right versions. They evolved over time but they were immediately useful to people.
With Sudowrite, the business is growing but there’s a part of me that thinks, “this isn’t it yet!” We haven’t found the right mix of features or the right positioning yet.
Some stuff is still up in the air.
That’s an uncomfortable place to inhabit.
But that discomfort also drives the financial decisions. We don’t want to hire 20 people, and have only a year to find product-market fit. Then not find it and disappear. Or perhaps we find it, but then the economy goes burst.
So closely monitoring our burn is really important. This goes both ways. Underspending is just as dangerous.
We’ve done a decent job of creating a culture of frugality. Not to the point where people are missing or lacking things. Instead it’s awareness that “hey, the mission here is to build a company that’s going to be around for a long time and be self-sustaining.”
We also believe that we can get there with a very small number of people.
Honestly, it’s more fun building with a small team anyway. Because then everyone is doing. There’s less overhead. There’s less personal conflict. There’s less management.
I actually feel that our choice early on to grow slow and steady has probably been the biggest driver in keeping the burn low and extending our runway. Simply because we don’t have to try to double every month/quarter. (That said, we 5xed this year, and I’m pretty happy with that.)
Specifically in relation to GTM, community and content, it’s an area where we’ve put surprisingly little attention.
It’s kind of laughable that we’re at 500K ARR right now and we haven’t spent a dime on ads, on marketing, on influencers, on social media. Our Twitter has a few dozen tweets.
And it’s very much the opposite of Photojojo. Which was a content-first business. A newsletter at first. We grew an audience and then built something to sell to that audience that they really wanted. I probably should be taking more lessons from myself.
James’ startup, Parse, also did a lot of content marketing that helped them get a lot of their early users.
This is funny because our audience is writers. So you’d expect that content would be one of the first things we do, because we have access to so many writers and because we are writers ourselves. But we’ve mostly been focused on building.
We set up a blog recently and have gotten some of our best users to help write it.
On the community side, we have a pretty active Slack community around the product. For people to exchange tips and help each other. There’s a lot more we could be doing there, too.
Potentially we could be doing something similar to what you’re doing with Relay where we create a vibrant community around AI writing, regardless of the tool you’re using and we let writers who’re interested in this new type of writing come and talk about it.
So we could follow your lead there.
— MagicBell’s co-founder, Hana Mohan, on “thinking of bootstrapping as a tool, not a way of life”
— YouCanBookMe’s co-founder, Bridget Harris, on the still-persistent, often glorified demands of bootstrapping
— Productboard’s co-founder, Hubert Palan, on why the market “opportunity gave [them] the confidence to raise VC funding”