4 ms·
I had the idea for the feature that would eventually turn into Unwaffle [1] back in July. It was initially just going to be an internal tool to track trial use
by jasonkester 10y ago
I had the idea for the feature that would eventually turn into Unwaffle [1] back in July. It was initially just going to be an internal tool to track trial user actions for S3stat so that I could detect trends to predict who would convert on their own and who would need intervention.
But it quickly became apparent how useful it would be as a product, so I built it that way from day one, with its own domain and API for collecting data. The idea was to build the pieces I needed to do my thing, and if it didn't work amazingly well, to shoot it in the head and move on.
So:
- Idea in July
- Prototype in August
- Decided to move forward in September
- Public site live in September
- Fully functional by around October
- Some minor promotion in November
- Just went live with real customers in December
No paying users yet (and no means of processing payment). With luck, and assuming the first round of trials go smoothly, it should get a proper "Launch" in January.
Naturally, feedback is always welcome.
[1] https://unwaffle.com https://unwaffle.com
- umut 10y agoCongrats! Great problem to solve, would you mind sharing some architectural decisions or technical details about how you plan to tackle scalability?
- joss82 10y agoAwesome! Could you elaborate a bit on the "minor promotion" part? We are at the "fully functional" step and wondering how to approach the marketing part. That'll make a great blog post, i guess we are not alone in that situation. Many thanks. We are http://www.parseur.com http://www.parseur.com by the way.