AL #032: Building BasePaint: The W1NTΞR Interview

How BasePaint built a one-of-a-kind collaborative painting protocol in public.

Just March of last year W1NTΞR kicked off a small announcement thread, setting out to build several web3 projects in the open:

Just a year later, he's already found success.

His latest project BasePaint is one of the most interesting and well executed NFT projects in this space.

As a fellow member of Solidity Guild he was kind enough to answer some of my questions regarding the story behind BasePaint, thoughts about building in public as a strategy and even some of the mechanism design details involved.

BasePaint, emerging during the “NFT bear market” is proof that truly innovative products will find users and thrive.

I hope it inspires you to think more boldly, creatively and collaboratively about your next project.

Let's dive in.

The Interview

Q: How would you describe BasePaint?

BasePaint is what I believe web3 should have been all about. Technically it’s a collaborative art project, where people from all over the world paint together. Broadly, I think of BasePaint as a living art form, a system with proper incentives that can function autonomously.

The BasePaint home page.

Q: I noticed that you're operating at the intersection of the web3 and “build in public” communities. Can you talk about your career prior, making the leap to building in public and how that led you to BasePaint?

I’m a software engineer obsessed with my craft. I wrote my first program 20 years ago before I owned a computer. Spent many years at Facebook, worked at big enterprises and tiny startups.

I love building tech products on the edge of tech and user experience.

The first time I looked at an Ethereum smart contract, I found it mind-blowing. I also couldn’t believe so few people actually see the beauty and the potential of this tech.

I created the @w1nt3r_eth Twitter account to share my excitement, things I learned, things I was working on. I found this process of building in public very rewarding. Trough this journey I met a lot of extremely smart and genuinely cool people. One of them is Zach, co-creator of BasePaint. Zach is an awesome designer and product thinker who shares my excitement for web3 tech.

Q: Collaborative art is inherently unpredictable. BasePaint seems to produce really high quality art. Why do you think the quality has been so high? How do the ideas of brushes, chat and themes contribute to it?

Our secret to high quality art is a delicate balance between professional pixel artists and people who are new to art. From when we brainstormed the project, Zach and I had a hunch that we’d have to have a back pressure system. If we open it up to absolutely everyone then it's likely going to be overwhelmed by low-effort contributions. If we make it paid, it’d also attract the wrong kind of people with the wrong motivations.

I regret making brushes tradeable. We were very short on time trying to ship this project and it slipped my mind. Trading causes a lot of suboptimal behaviors. E.g. it’s now somewhat profitable to operate a bot that mints a brush and sell the brush on OpenSea for a quick profit. We keep fighting these bots with tricks, but it’s a battle that will never end.

Keeping the quality high requires constant work. We have a moderation tool that helps us flag bad actors. 20 minutes before the canvas becomes mineable, we lock it only for people with artists brush.

The chat was funny. Zach mentioned it’d be cool to have a chat and live cursors, but that was a P2 (priority 2, which means almost never gets done). I took it as a challenge and tried a few things.

In the end I’m super happy that we did include the chat, it definitely transformed the whole experience.

Q: The tech nostalgia and pop culture themes and colors seem to provide just the right amount of constraints to create deeply interesting pieces. What's the story behind the voting system? How do you think this protocol generalizes?

Initially we didn’t have “themes”. However during the beta testing a few days prior, it became obvious that just a canvas is not enough, it’s unclear what to do. We needed a nudge. So we added a small prompt at the top, our first one was “Onchain Summer” (inspired by Base launch). On Day 2 we figured that reusing the same colors will make things confusing, so we also changed up the palette. For the next 30 or so days we curated a list of theme names and palettes ourselves. Some were fun, some ended up very challenging.

We definitely wanted to transition to a state where the owners of the artwork could influence the themes. We started with a Prop House, but that felt very heavyweight, a round lasts 7 days and people needed to remember to come back. But overall our users like the idea of being able to suggest the themes. So we’ve built out a simplified version of that right into BasePaint. Now there’s a reason for everyone to come back every day.

After running voting for a bit, we noticed the themes were not great — too many public figures, memes, inappropriate things. We’ve added a way for artists to flag themes, and also hid the votes for the first 12 hours of voting. This helped a lot, the recent few weeks worth of themes are really nice and surprised us with creativity.

Q: Multi-sided marketplaces typically suffer from the “cold start problem” (Andrew Chen). You've talked about carefully selecting the early users. How did you overcome the cold start problem for BasePaint?

On the artists side, we searched the artists among our friends and friends-of-friends. I also tapped into some Discord servers I was part of. Most of these were pretty dead (crypto winter is no joke), but the users who remained active were actually the perfect audience — committed to the long run.

To find the first set of artists, we tapped into our network of friends and communities. This, again, was when my “building in public” helped a ton — people knew me and had some base level of trust. We’ve also cold reached out to many pixel artists on Twitter. Most of my DMs were unanswered (there’s just too much scam in web3). However, the few people who did reply became the legends and helped shape BasePaint into what it is right now.

As for users and collectors — we launched with partnership support from Base. I reached out to Jesse Pollak early on when they announced the launch of Base and the call for dApp developers. So we knew we’ll at least get an initial publicity boost. We got featured on their websites, Brian Armstrong minted Day 1 canvas and posted about it, which attracted a lot of attention.

What was surprising is that after the initial boost passed, we did see organic growth.

Q: BasePaint gained attention quickly. You did interesting things like sharing videos of how pieces were created. Can you talk about some of the surprising distribution tactics that worked well?

I think the artwork lands itself very well to social media. It’s visual, engaging, different every time, but always “consistent” in a way that you can recognize BasePaint painting.

Early on many artists complained that their art was painted over, so a community member made a script that replays the history of each artwork piece and creates a video. These videos became a huge hit! We now automatically post them every day on our Twitter.

What was also interesting and unexpected is that some people have started (ab)using these timelapses to create videos! You can see it on Day #9.

where an eye is traveling across canvas, but you can only spot it on the video.

Another really cool thing we see that helps us grow is partnerships. On Day #9 we did a “Pixelmon Kevin” theme, which boosted our mints from 300 range to 700 range. Our XCOPY theme was a hit too, ended up on Times Square.

We also “permissionlessly” set a Jack Butcher theme, and Jack appreciated it and shared with his community (which is also very aligned with the values of BasePaint). Our latest collaboration with Lens brought a lot of attention from Lenster ecosystem.

Q: How have you handled community management for both the artist, curator and collector side?

We don’t have an official Discord, which kind of helps a lot. Community management is a full time job, and we are glad our community sort of manages themselves.

We have ephemeral chat on the side, which is where live discussions happen. It’s very cool to see people onboarding each other, answering questions, coming up with ideas, etc. It also makes BasePaint feel like a place, cozy, fun, creative.

On principle, we don’t overpromise, we don’t want to set any expectations and create an opportunity for people to get disappointed.

We love the ideas of decentralization and voting, but we also don’t want to do it too early. Many people have asked for treasury and DAO, and I don’t think it’s gonna help the project at this point. It’s still fragile and things can break.

Q: What advice would you tell readers who are looking to build in public in web3?

It’s freaking hard to break through the “nobody cares” wall. I’ve tweeted for at least 6 months on my w1nt3r_eth account until people started noticing. But I did it because it was really fun, I was very excited about the tech, so I didn’t care much for not having a big audience.

(My little secret for tweeting is Typefully. When I feel inspired to write something, I’d make a few tweets there and schedule them. At times I had an outgoing queue of 20+ days worth of content.)

Build a v1, or “first season”, or whatever helps you overcome the perfectionism and ship something.

Reduce your target audience. Intuitively everyone wants as many users as possible. But it’s really a bad strategy. Generic products are not appealing to anyone.

Q: How did you come up with BasePaint?

I generally don’t believe in ideas being “created”. I think ideas are like a field, constantly changing and interacting with each other. I can trace back some inspirations to Reddit’s The Place, my own project SNOW and

The Million Dollars Homepage.

What’s really awesome is when you combine your ideas field with someone else’s. In BasePaint case, chatting to Zach has allowed for both of us to shape our idea fields into a product that works.

Q: In terms of the community section, are there any behaviors you are observing, e.g., people forming Telegram groups or ways community standards get enforced in chat?

We don’t have Discord because we don’t have the time to monitor and curate it. We have also been burned by Discord mania of 2021 where each project was promoting their private chat channels, overpromises, drama, etc.

I think there are private sub communities of BasePaint fans, e.g. we know there’s one in Turkey where people coordinate and discuss things, but we are not involved in managing them.

We’ve also seen things like BasePaint Fan Club party on Party DAO. I really like this idea, but it’s not very active. E.g. we gifted the community a silver brush but they haven’t managed to use it yet. Coordinating humans is a very hard problem.

Re: autopilot, we are long ways off from the dream of self-government. Even the most successful and the wildest experiments in crypto like Nouns struggle with governance. Our dream is a self-sustained BasePaint ecosystem where the creators of it can step down, but for now it’s clear we still have a lot of work to do.

I’m hoping BasePaint will inspire a lot of copy-cats. E.g. we are seeing Daily Jam that has similar mechanics. I think the BasePaint model where a group of people creates something every day and profits from it every day is really cool and I’m hoping to see more of these.

Some more information about BasePaint can be found on our Optimism retroactive public goods funding page.