a single source, nothing corroborates it

Leave the split tests until later

Copy tweaks and page optimisation are mechanical, they stay available, and somebody else can do them, which is why they belong after the thing that actually decides the business. Hours of iterating toward a fifteen percent lift compete directly with the same hours spent trying for two or three times on the one number that matters. Do the big one while it is still yours to do.

The method

1. 01

   Postpone split tests, ad copy tweaks and page optimisation, because that work is mechanical, stays available later, and can be handed to someone else.

What it returned

• 15%

  He weighs hours of iterating toward a 15% lift against spending the same effort trying for two or three times on the one number that matters.

Sources

• The full story of “the one important thing” for startupsJason Cohen · longform.asmartbear.com · The full story of “the one important thing” for startups · 2011-10-03