Designing for experiments & A/B @etsy Dan McKinley #warmgun
Do one very minimal change at a time.
Measure, understand and move on.
Plan to be wrong.
Prefer incremental design changes instead whole redesigns.
Why we experiment and measure?
For surprises like:
- dont open new tabs - people not used to it
- Infinity scroll good or bad? - not good for etsy
- Auto suggest in search increased sales by 4%