I’m Mac Budkowski and I’m a co-founder of Kiwi - a project similar to Hacker News, but focused on Ethereum. So we deliver the feed that covers close to all interesting Ethereum-related content that can be found on the Internet.
In the last 2 years, we spent a lot of time figuring out how to create an environment that makes users engaged, so here are some thoughts that I’d like to share with the HackMD community.
We built Kiwi because crypto moves very fast and quality content is scattered across Twitter, Telegram groupchats, Discord channels, blogs and newsletters, so it’s hard to keep up. At the same time, Hacker News has been very hostile towards crypto, so if we wanted to fix this problem, we had to build an alternative.
To ensure we don’t turn into Hacker News, from Day 1 we are open source and store all our data on top of a P2P protocol. So if someone doesn’t like our app, algorithm or moderation, they can fork us and build their own app, accessing all Kiwi links, upvotes and comments.
So far we have curated over 15,000 links and shared over 6,000 comments, and have about 1,000+ monthly users who check our app. We also have a community of a few hundreds devs, founders and people who work in Ethereum who hang out on our groupchat.
Right now, Kiwi focuses on crypto tech, products, and culture, but the architecture is designed to scale across any niche that needs high-signal curation. So this is how I got into communities.
We have both the web app at https://news.kiwistand.com/ As well as iOS (TestFlight) app: https://testflight.apple.com/join/6jyvYECH
What community-building means to me
I think real progress never starts in the main hall. It starts with 5 people eating conference lunch together, a niche subreddit or a 50‑person Telegram group grinding on the same weird problem at 3 a.m.
If we look at the history, this was the case with French impressionists, Project Manhattan, or Cypherpunks. When we put niche-obsessed people in the right environment, they move the whole domains forwards.
So building infrastructure that helps niche-obsessed people stay productive feels like a practical way to contribute to the development of Ethereum, and - in the future - other projects.
I have been thinking about this subject for some time and even wrote about “Niche Internet” and “Why niches matter?"
Successful community building
I think it’s key to create a space where people show up not just to passively consume, but also to contribute.
In our case it’s when users actually submit and discuss links, shaping our feed. Or they give us ideas on how to improve the product. Or they spread the word about the project - either by putting our stickers on their laptops or by mentioning us in the discussions on social media.
I am also a very IRL person, so for me good IRL meetups are an important indicator that something is working. Since our community is very global, we meet mostly at ETH conferences. So last year we met in Bangkok in a Korean restaurant, and an hour before the meetup I felt like everything was ready. But then it started raining.
Not our European drizzle - a real, hard tropical rain that made the already tough city traffic crazy. I was worried that people wouldn’t make it. But - a bit late and soaking wet - they made it and we had a great evening.
When people put in the work to meet IRL, it’s an honest signal that they really care.
Fun fact, Tom Preston-Werner, GitHub Co-Founder, said that in the early days he was travelling the world and having beers with all developers he could find to build the initial community. So if IRL is so important for such a digital-first product as GitHub, it’s probably important for everyone.
Getting the community to engage
When we first launched Kiwi, we had close to zero content. This meant we couldn’t create a valuable feed, and since we couldn’t create a valuable feed, we couldn’t get users. A classic chicken-egg problem.
So we started asking friends to send us 3 links they find interesting, and we put them on our website. So this was Kiwi v0.1 - just three links a day, some about crypto, some not. We called it “Editor’s Picks”.
Then a user named freeatnet messaged Tim and started discussing our tech setup. Soon after - although we were a very young project with close to no users - he built his own app on top of our P2P protocol, with a different UI and algorithm.
It seemed like the way we built Kiwi - open source, very transparent - was what has drawn him, not our DAUs. Later on we received many contributions from our community.
Misha started submitting great links every day, and he still does to this day. Our friend’s wife - Caroline - designed our logo and stickers. Another user - Matallo - built the first Kiwi search engine. Once he had to stop maintaining it because of other obligations, another user - Rainer - built a v2 of search engine that’s still running.
We had some users helping us with design, some - like Brais - printed stickers at ETH Barcelona and just gave them out to people. There was also Chris who helped us get some grants to get things going. There’s also Pugson - who runs ENSData that we use in our app - that recently donated over $1,000 to our project.
And there are countless others who gave us feedback, supported us, and helped us remember why we do it. Being a founder is a lonely job, and since it’s just the two of us, having such a community support makes a huge difference.
Overcoming challenges in a community
I think the biggest challenge is staying focused without killing openness.
When you build an app like Kiwi, it’s tempting to let the community post whatever they want. It’s a community project, right? But if you’re not careful, the feed quickly drifts - someone shares cat videos, it gets upvoted, and suddenly you’re Hacker News for memes.
We saw this early on and decided to actively change the course. People started asking: “Shouldn’t the community decide what content we want?”
That was a fair question but we knew that if Kiwi didn’t stay focused on high-quality crypto links, we’d lose the signal that made it valuable in the first place. So we had some tough conversations and introduced clear content guidelines. Some users left. But the ones who stayed helped make Kiwi what it is today.
Also, users always have many product ideas. If they’re active community members, they have even more. But our capacity to ship is limited, so often we have to tell people that we can’t deliver some feature or improvement they care about, and it’s very hard.
Collaborating within a community
If you look at an average online community, it’s very different from an offline one. There are people from many countries, living in different timezones, having different goals and experiences.
So there’s higher variance when it comes to people, and it’s also harder to communicate clearly because we just read text instead of having IRL conversations. That’s why we try to communicate a lot.
Last time we changed the philosophy of our feed, Tim wrote a long post in our Telegram channel explaining our thinking behind it, and discussed it with the power users.
We also share our doubts and mistakes, because we believe in building in public and being transparent about our journey.
We are also both pretty straightforward, so we often discuss things that are considered taboo, which helps to strengthen the bonds between us and the users.
Last time we even created this Signal group where messages disappeared after 30 seconds, because the discussed subject was hot and we wanted people to feel safe to share their takes.
How I use HackMD
We use HackMD almost all the time when we apply for grants or write bigger blog posts. Being able to just use Markdown is the feature that sold the product to me.
Would be great to see HackMD become even more multiplayer-focused. At the moment we have multiplayer features for our notes - we can invite different people to share comments, run our own team and so on.
And I think we could take it one step forward. At the moment I see people’s HackMD links only when they share them on Twitter or groupchats. But what if you ran a Discovery Feed where I could find other people’s notes based on my interest?
I could follow people’s half-baked ideas, add my two cents to someone’s draft, and meet people through that channel. I think that could be really interesting.
How to strengthen your community
Communities don’t live in a vacuum.
Just like products need to address users’ needs or otherwise will be ignored, so do communities.
If the financial system was working well, no one would be interested in Bitcoin. But it did not work well, so people had a natural need to look for an alternative. And working on Bitcoin was a way to channel your frustration into something productive. Same with Microsoft and Linux or Internet Explorer and Mozilla.
So before you start building a community, it’s good to ask yourself: what kind of problem do I want to solve for these people? Last year I did a short talk about our lessons learned, you can listen to it here.
I think it’s easy to focus on quantitative metrics here: how many people do you have in the community, how many posts do they share, how many comments do we get and so on.
While important, I think it’s crucial to remember about qualitative metrics. Most projects would rather have a community with John Carmack than 100 junior devs. Or have 10 people sharing thoughtful, Hacker News-level comments than 100 just saying ‘hello’. Or have 3 code contributors than 30 yappers. So although the numbers are important to us, we look at the quality a lot.
What I love about HackMD
The way I look at it is that if the GitHub team wanted to create Google Docs, they’d create HackMD. So just like in Google Docs, collaboration is the foundation of this product, not an afterthought feature that’s buried somewhere deep in the UI. And the whole product has a strong hacky vibe that I really enjoy, and I think it makes it exciting for more tech-oriented communities.
If I were to name specific features, I’d say setting up teams, following people’s inputs, syncing with GitHub, and - what I particularly like - letting them sign in via Ethereum wallets, is great.