How do you organize and prioritize growth experiments?

[krnauau]

I've seen theAirtable from Airable's head of growth, but still not sure how others are doing it.

How do you decide when an experiment is worth running?

How do you decide whether to test with a "quick hack" or build for production? How do you handle when a quick hack works, but not enough to do the work to make it work in production?

How do you re-use experiments in other experiments? E.g. use email customizations in web personalization. Is that just rebuilt from scratch in each tool?

 

[whattheheartwants]

@krnauau Well, the first thing is you don't need to build such a complicated base to start with. You can make a simple one.

Date question: The start date and end date of different experiments differ.

How to Prioritise? You can prioritize experiments based on ease of execution and the impact it would make.

For example. for a particular experiment, I would need 5 hours from developers and Impact is medium Vs I have another experiment where I don't need developers' help and the Impact is medium.

In this case, you can prioritize the latter. Likewise.

I might make a simple Airtable base for experimentation. If you want, I can share it with you,

 

[krnauau]

@whattheheartwants Thank you. Sometimes I’ve struggled when the impact is potentially high but untested. So we reduce the effort by building as a hack. It’s better than not testing, but sometimes the hack becomes production and requires more maintenance. Do you ever see that? Are there any patterns to design quickly but prepare for scale?

 

[benjy_lastsonofthewest]

@krnauau Assessing the difficulty and impact of different experiments is a good place to start but the results of previous experiments as well as customer interviews and/or user testing should help inform your decisions as well.