3 ms·
I think if frame it as, “each 100ms you shave off a critical consumer shopping checkout flow page increases conversions by X%” it starts to seem more defensible
by rcoder 4y ago
I think if frame it as, “each 100ms you shave off a critical consumer shopping checkout flow page increases conversions by X%” it starts to seem more defensible.
Agreed that in the general case 100ms to full paint (with interactivity not far behind) is really good.
- scoopertrooper 4y agoDoesn't really matter if the relationship isn't linear like that. I suspect it's more likely conversions start to drop off after certain thresholds are met e.g. 3 seconds causes conversions to drop by 5%, 20 seconds causes them to drop by 30%. Just averaging out those numbers could result in engineering time being wasted chasing rapidly diminishing returns, if the site is already below one of those thresholds.