A three step process for documenting, analyzing and synthesizing your whiteboard session.
This is part 2 of my discovery research series of articles… click here for part 1.
Everproof’s technique for documenting a premortem
- List (5 mins) — Copy all your post-it notes into a spreadsheet, you can do this in a few minutes if you type quickly and use spell check to fix any errors at the end. Use one column to list all the assumptions and another column to list all the reasons for failure. Make sure you keep them in groups in the spreadsheet too.
- Clean (20 minutes) — Cleanup each list to remove duplicates and add some consistency to the syntax.
For example, these three cards are all referring to a similar thing:
- TFB — Our terminals cost too much
- TFB — Our solution was too expensive to manufacture
- TFB — The unit economics did not stack up
I would merge those into a single one, and probably rewrite it as: “TFB we couldn’t manufacture a device at a low enough price point.” Again, make sure you keep them in groups in the spreadsheet too.
2. Rank (20 minutes) — Get all participants to rank the list of assumptions and reasons for failure. The criteria for ranking should be “impact on the business”, with a 3 for high impact, 2 for medium impact and 1 for low impact. Make sure you do this task in isolation to avoid groupthink. Add a new column for each person’s ranking to your spreadsheet.
3. Discuss (20 minutes) — Next do some simple analysis on the data. Create a total, median, variance and mean for each row. Play around with sorting the columns and having a discussion about each rows impact on the project. Pay attention to areas where there is a higher variance, it’s healthy to disagree but you should try to understand why. You want to end up with a list sorted by total votes, these are the most important assumptions to challenge and reasons for failure to protect against.
Next we’ll use research to figure out if our assumptions are valid or invalid. We’ll also create a failure prevention plan using the reasons for failure. I’m currently writing those posts and will publish them in October.