Getting feedback as early as possible
I decided to make a game that helps you learn both geography and historic events. I started with the geography angle after falling in love with Geoguessr, and tried looking for things where the market isn't saturated already, which is how I landed on historic events.
While I'm most interested in eventually making a mobile app out of this idea, I decided to build a web version first to test the concept out. First on friends and family, and then more widely. So I'm building Wherio - it's a daily game where you get a historical fact and have to match it to the country on a map where it happened. There's 5 rounds, each getting progressively harder.
I started sharing with just my immediate family and got some useful feedback; too short! So I changed the game from 3 rounds to 5. I'm now seeing an average of ~3 minutes for puzzle completion which feels about right for a daily game.
Then onto friends; and the feedback is still rolling in but overwhelmingly...it's too hard. Good to capture this one before sharing more widely so I can adjust the leveling.
Then I went online and started trying to find ways to get strangers to play. I am learning that while I may not be very good yet at soliciting feedback on Reddit, those sorts of posts are driving users to play, which is really helpful. The more folks I can get to play, the better chance I have of getting to an answer on: is this worth continuing to build?
Very early data shows some promise:
Folks are at least completing the game about 70% of the time. Granted some of these people are friends and family, so I expect the realistic number to be lower.
The most important piece though is still to be determined; do people return? Ideally I can get at least 25% of users to return the next day, and 5% by day 30. I'm setting some rough goals based on iOS game statistics, which may not translate directly to web games but is at least a similar proxy.
The hardest thing right now is to avoid biasing on the feedback that I want. Ie, I'd love to keep building this. But if enough people aren't interested, then it's probably not worthwhile. So that's the feedback I'm after right now: do strangers use it? Do they keep using it? Where do they fall off?