After a middling launch last week, I came away with a few dozen trial signups and no paying users but some excellent feedback. Saascast started life as an internal tool and I still use it every day in that capacity, so even if my hopes of a splashily successful launch were dashed I'm thrilled to have gotten some great feedback to help me improve the app.
The best of these was a HN comment asking if the forecasts were the output of an ARMA model. They weren't -- when I launched on HN the forecasts were the output of a hybrid weighted regression -- but now that I've had a few days to do some research I was able to add in a much more powerful forecasting model into Saascast. Thanks, random HN poster, for a pointing me in the direction of some truly excellent mathematics.
Very HN of you, I suppose. But cheers nonetheless! 馃