This part closes the Experimentation Cycle.
Sharing experiments helps others to profit from your learnings. Sharing does not (only) mean to tell your next colleagues over lunch in the cafeteria how big your uplifts are (although you should do that, of course).
<aside> 💡 Share = Document + Communicate your experiment
</aside>
Documentation sucks. Especially if somebody else (or your younger self) didn't do it and forces you to lose time searching for experiment results. Hands up everybody who had that moment: "Hm, didn't we test something like that last year? I would have to look it up. Shit, where was that again??? Damn, I find the figures, but not the layout. And these figures here.. were they the final ones...? Were the results even significant? I can't remember!"
Documentation can be more fun if you ...
We all know that we should document our outcomes regularly and centrally for the greater good. What makes it so hard, is when documentation comes as an extra step at the end of a process. If documentation and collaboration come in the same tool at the same step, documentation is easier. A combination of Confluence and Slack has proven to work here.
Over time, I developed a routine in what is worth documenting. This is reflected in the Experiments Template in Confluence. It guides you along with the steps of experimentation, helping you collaborate and document.
It is very rare that you work on an experiment alone. There are always colleagues who will contribute to the test or the experiment. Define in advance, who is responsible for documentation - and hold that person accountable for it. This does not necessarily need to be the experiment owner, although it might come naturally for him to do so.