I generally agree (with caveats and limitations of this hypothesis as well!). I'd add two things. The first is that Warpcast is a very good crypto app. This means that it might be much harder to take people off Warpcast than to take them off X, Reddit, or Discord. It's like asking a person to try out some new food when they're already satisfied with their meal vs. asking someone who doesn't like what they eat (as many people don't really like web2 socials AD 2024). The second is that "being active on..." is generally not a good strategy. Whenever we use X, Reddit, or Farcaster for user acquisition, our actions should be focused and have clear goals. In Farcaster's case, pulling it off is harder, as we also use it for socializing and having fun. So it's very easy to tell ourselves, "Oh, I'm there for work," whereas in reality, it's just posting random stuff instead of deliberately planned posts that, in the short- or long-run, will turn into converting users to Kiwi.
Interesting how the data debunks the usefulness of Farcaster for user acquisition. I would have bet otherwise. Still stand behind FC users being higher quality crypto users though. On the bots, I'd disagree that getting rid of them would increase app usage in a meaningful way. The bots are good for coming across new/hot submissions - I still open the PWA for dedicated exploration and reading. in fact, the bot can lead to ad hoc reading. I doubt the bots are hurting DAU. But would be good to see what the data says.
Great data point @sahilk. So you see the bot posting in your Warpcast feed and it makes you browse Kiwi more? If so, how can we make the bot even more useful for you?
It was also counter-intuitive for me, and I agree with the limitations mentioned by Tim - that the data from Google Analytics is imperfect. Probably, the real numbers are higher. Good to see a vote for bots because I also like them. I think bots are tools for low-key retention. The way I look at it: high-key retention = PWA app, mid-key retention = newsletter, low-key retention = bots. So, if someone likes Kiwi a bit but is not convinced enough to download PWA or subscribe to a newsletter, they can follow our bots and, from time to time, serendipitously come across our links in their feed.
Just curious, among all the kiwi link submitter, what percentage are from farcaster?
Just a guess, but I'd say that the overwhelming majority of people who submitted a link to Kiwi News do also have a Farcaster account. So yeah, that is a pretty good caveat to the blog post I wrote because it rejects the finding in the Google Analytics data. A few more observations about the bots: - It can also be that Kiwi users deliberately do NOT "like" @kiwi or @kiwinewshq's tweets/casts because they click through to the Kiwi News to upvote the link - In general, when browsing links using the bots, the intention is to click through to Kiwi News (at least) because that's where the original link is, which an active user would seek out.
@timdaub.eth The bot for me is largely unintentional discovery. If the link looks interesting enough, I click and get reading. Since I'm already on the Kiwi page, I might read a couple more too, depending on the time available. What also helps is the bot telling me who made the submission. Seeing names I already associate with quality links increases the likelihood of me clicking. Btw just noticed, I don't follow the bot. It's you and Mac casting Kiwi links / recasting the bot that does the same job for me. The utility and behaviour still hold true though. Maybe highlighting the bot's role in discovery when I minted could've pushed me to follow.