By Teresa Torres · producttalk.org · @ttorres on X · LinkedIn
Teresa Torres argues that predicting the expected impact of each user story before building it serves two purposes: it helps prioritize the backlog, and it surfaces the hidden assumptions behind what a team is building. She acknowledges predictions will often be wrong, but says the gaps between prediction and reality are where the real learning happens. Using an event-creation website as an example, she breaks each story's impact into adoption rate and per-user effect, then shows how low adoption can outweigh a strong feature effect. The method's value lies in prompting teams to ask why results diverged from expectations, and then using those answers to sharpen future predictions.
01Key takeaways
- Estimate expected impact for each backlog story to inform prioritization alongside other factors.
- Break impact into adoption rate and per-user effect so each assumption can be tested separately.
- Adoption often matters more than effect size: a widely used modest feature can outperform a strong but rarely used one.
- Treat large gaps between prediction and result as the most valuable learning opportunities.
- Investigate why adoption differs from expectations before optimizing, since privacy concerns cannot be fixed by tuning.
02Key sections
- Prioritizing with expected impact
- Estimating what each story should deliver gives teams useful information for ranking backlog items, though it is one factor among many.
- Predictions expose assumptions
- Forecasts are almost always inaccurate, but writing them down before release creates a baseline to compare against actual results.
- Breaking impact into components
- Overall impact is modeled as adoption rate multiplied by the per-user effect, for each of the two example features.
- Learning from the gaps
- Large differences between predicted and actual outcomes, such as low adoption, prompt questions about discoverability, usability, or privacy that guide better decisions.
- Iterate on predictions
- Teams should record predictions, analyze the gaps, and use the lessons to improve the next round of estimates.
03From the post
“On Tuesday, we looked at how to measure the impact of each user story, looking at both measuring the effectiveness of the mechanics of the feature and also the impact of the story on our overall product goals. Today, we'll look at how to predict the impact of a”
Summary and takeaways written by PM Atlas; quotes are short excerpts. © the original author.