How do you say PMF in Portuguese? After long months of coding remotely with X2 from my parent's apartment, we shipped a testable PoC. It was time for a confrontation with the first users. Factchain is available for Warpcast on Base mainnet (with ridiculous stake amounts) through a frame and for X (ex-Twitter) directly in the browser with the extension. We announced on Warpcast first to receive feedback from a benevolent community. X1 took over the communication. The general atmosphere on Farcaster (the social protocol behind Warpcast) is known to be more indulgent than on Crypto Twitter, and even though it's too early to have misinformation posts, the potential early fact-checkers were here, living at the intersection of crypto and information. As a reminder, Factchain brings cross-social onchain community notes. The community welcomed the project with warm interest, and we even saw a bunch of notes starting to appear under Tweets without having shared a single communication post on Twitter. In my eyes, it was enough traction to start talking to VCs. This was all very new to me. X1 and X2 agreed to join the venture full-time if I managed to raise enough capital to match their current salary. I put on my CEO hat and began the mission.
How to fail a fundraising?
There are many excellent books on how to succeed in fundraising. I thought I could contribute to this large body of literature by talking about failure, too often overshadowed by industry fairy tales.
Here are my top tips for failing to raise funds. 100%.
Having "co-founders" working a full-time job. This take is obvious, but entrepreneurs sometimes don't have all the options available. It creates a cap table that no VC can comprehend. It requires a bigger round. Production stops because you're the only one full-time and busy looking for investors all day. VCs are like sharks; they smell blood miles away.
A GTM strategy that is too disconnected from the end product. How does Factchain Community enable Factchain Media? We saw the fact-checkers community as the only solution for mainstream socials to integrate Factchain Media. Were we right? I don't know, but it creates too many "what ifs?". Bootstrapping a community is already a full-time job; I think VCs saw it more like two different products and missed the connection. A GTM strategy should de-risk your project, not add more interrogations.
Aim for a large financial round too early. I had many reasons to justify expenses. Among other significant developments, we needed a better note ranking algorithm than Twitter, and the Factchain PoC only computes an average of all ratings. I was disconnected from reality: We hadn't proven the beginning of a PMF, and the tech requirements were probably already too big. Better start small with BAs before pitching to top-tier VCs.
VCs feedback
Even if the long-term goal is to go mainstream, Factchain is too industry-specific for generalist VCs to lead a pre-seed round, but I had the chance to be introduced to one of the most successful crypto VCs. After much discussion on the project, they couldn't believe in our ability to (i) bootstrap an audience and (ii) make it sustainable. In yesterday's post, I promised to share why the notes and rating stakes were probably a bad idea. A Lannister always pays his debts.
The note and rating stakes didn't help at all to reassure VCs on their aforementioned worries. Remember, we developed Factchain with the "Put your ETH where your mouth is" narrative of X1. Users would pay some ETH to submit notes and ratings and get rewarded or slashed based on the quality of their contributions. These financial stakes became our Achilles' heel. It raised a plethora of questions:
How do you pay for the winners? Factchainers are definitely not playing a zero-sum game.
How do you mitigate security risks? Stakes are locked in the Smart Contract for the duration of the "note auction" (i.e., the period during ratings are open)
How do you stop malicious actors from rating their own notes and always win the auction?
On the other hand, what do we really gain from stakes? The main answer is incentives, but I'm not even sure it's a valid one. Who wants to fact-check information for a few bucks? People interested in the Truth do it for other motivations than money. The cultural NFT reward would've probably been enough for them to enroll. Bragging about being right is a far more joyful incentive than making a few WEI out of it, and users could've been rewarded in an Airdrop anyway. Have you never heard of the FACT$ token? No, and that’s alright: it never existed.
We saw yesterday how the stakes could've secured the system by increasing the costs of network collusion. Intuitively, the more users involved, the more money a malicious actor needs to spread a single misinformation note. This intuition is wrong. The median rating number on X community notes is only 72 for 400k involved users. Submitting more ratings on your own notes will always be affordable unless requiring a dissuasive stake amount.
We should've explored other solutions, like reputation scores.
In most Web2 forums, like Stack Overflow, you may not be able to post a question immediately after your registration. The site has certain restrictions to prevent spam and ensure the quality of the content. Users must answer a few questions first or wait a while.
We could've replicated this reputation scores system onchain to gate note creation by Ethereum addresses with already a few correct ratings under their belt. And restrict the number of notes new addresses can rate in a given period.
A reputation score-based system would've drastically reduced the cost of the Factchain launch on mainnet: no winner to pay, fewer security risks to cover.
Retrospection
There is a demonstrated growing interest in the X Community Notes feature. I've written two Python scripts to extract notes stats and ratings stats from the X public dataset, and their output is impressive: 3x more users and notes over the last 6 months. The use of browser extensions and frames could effectively redirect this traffic onchain, creating opportunities to build a community of fact-checkers. This potential is worth further exploration.
Per Ardua Ad Astra. (3/3)
