just stumbled onto a breakdown about why media forecasting goes south. it is never actually about being bad at math, but more about how we predict the future. most of us have been in that spot where the plan promised a 4x roas by month two but then cpcs start climbing and ctr drops off a cliff. it feels like you are
failing at strategy just failing at the math part. i realized my previous models were wayyy too optimistic about stable costs.
it is usually an issue of ignoring external volatility . instead of relying on old data, we need to focus on
predictive accuracy rather than just chasing high numbers. does anyone else use a specific formula to account for seasonal spikes or do you just wing it?
>the real problem is the forecast itself.article:
https://neilpatel.com/blog/paid-media-forecasting/