<aside> 🚧 DISCLAIMER: The following standards and guidelines help you execute quantitative Delivery or Post-Validation Experiments mainly. For other validation methods, please contact the respective go-to person.
</aside>
Your goal is set, you picked the idea and decided on a suitable validation method to get you there. You are ready to start working on your experiment!
But hey?! What happens if you can think of more than one experiment? Fact is: it is very likely that you can think of more than one! Probably there are two or even more:
Probably, the majority of these experiments are variations of your initial experiment because you can't decide on the how. Structure them by differentiating what you will change where and to which user group to challenge your assumption.
For Example:
If your target outcome is "I want to achieve that my net revenue increases by 8% within the next two months (coming from 81.000 CHF, rising to monthly 87.500 CHF)",
your strongest opportunity would be to "increase the product usage"
and your prioritized idea would be "I assume I can reach that by enhancing the "Favoriten❤"-Function."
then your experiments could be: Experiment 1: I would animate the "Favoriten" Heart once for every user on the list page Experiment 2: I would add encouraging mouse-over texts on the "Favoriten" collection Experiment 3: I would change the heart to a "Reserve for me"-Button for every ad Experiment 4: ... well, I think you got it 🙂
So I recommend you to start with one experiment. Write down or scribble them all. You can use the Experiment Abstracts Template for it.
And again: you have to choose. Which experiment should you pick? Which experiment should you leave behind?
<aside> 💨 Pick the experiment which will deliver the fastest results.
</aside>
What good is an Experiment which needs 3 iterations of development and four iterations of run-time until it throws a result? You want to learn NOW! So, at this stage, it is all about speed.
Some rules of thumb how to generate speed while experimenting: