Content pfp
Content
@
0 reply
0 recast
2 reactions

〰️➰〰️➰〰️ pfp
〰️➰〰️➰〰️
@jklb.eth
At what point is an app “fast enough” and, you should stop looking for perf gains? Apart from biz metrics (read money from users) - how do you measure performance impact on users?
1 reply
0 recast
3 reactions

Apt pfp
Apt
@aptlab
Rendering time metrics are tricky because in actuality it’s really more about perception, in application, than true wall clock time. Is the A/B testing just showing that it’s slower to load, or that it’s definitively worse for end users? It’s said, humans can only perceive at a max rate of 13ms.
1 reply
0 recast
0 reaction

〰️➰〰️➰〰️ pfp
〰️➰〰️➰〰️
@jklb.eth
We’re actually getting push back that we need to make things more performant before A/B testing… “too much revenue on the line” kind of thing. We’re pushing to get them to test the assumption that “slower ALWAYS means lost value”, but this is a big machine…
0 reply
0 recast
0 reaction