Prioritize hypotheses based on ICE model
Prioritize hypotheses based on ICE model
1. Create a spreadsheet and list all your hypotheses in the first column.
2. Consult all relevant stakeholders to collate hypotheses for all upcoming experiments.
Use a tool like VWO, which has an in-built hypothesis repository.
3. Create columns for Impact, Confidence, and Ease in the spreadsheet.
Impact: how much the experiment is likely to positively affect the key metric. Confidence: the probability that the hypothesis is true. Ease: the amount of effort, in terms of time utilization, required to run an experiment.
4. Grade each hypothesis on a scale of 1 to 5 for impact, confidence, and ease, with 1 being the lowest and 5 being the highest chance of success.
Consult and discuss with various stakeholders, including experiment creators, product managers, developers, designers, etc. while grading impact, confidence, and ease.
5. Add up the impact, confidence, and ease scores for each of the listed hypotheses and divide the total by 3 to calculate an average prioritization score.
Add the score in the Total ICE score column.
6. Sort your hypotheses on the Total ICE score column or add a color gradient to show priorities.