Make business impact visible before we commit, not after. Three disciplines: a sensitivity model on live data, a baseline of past projects' impact, and a matrix to extrapolate the lift of a 1% move on a chosen KPI.
The PM's job is to make the impact visible up front, not to discover it after release.
Build a simple model based on the data captured online over the rolling 30 days. Use it to estimate the impact on revenue of a lift on a specific element — a lift on a business KPI, a lift in the volume of all or a specific user segment audience, a lift in a specific acquisition channel, or a lift in the activity volume in a country. This high-level estimation keeps the big picture in mind when looking at what we are doing. See attached document.
Combined with the sensitivity model, maintain a table of past work and the impact it has had: the Baseline Impact file. This file is also the visibility given by the Product Department to the other departments and the CEO.
| Project | Segment | KPI moved | Magnitude | Window | € impact |
|---|---|---|---|---|---|
| Cart quantity selector — mobile | Returning non-buyer | Cart→Checkout rate | +3.2 pts | 14 d | ~€11K / mo |
| Personalized HP carousel | Returning visitor | HP→PIP rate | +5.1 pts | 21 d | ~€18K / mo |
| Filter bottom-sheet (mobile) | All mobile | Catalog→PIP rate | +4.7 pts | 14 d | ~€9K / mo |
Build a matrix to calculate the theoretical revenue lift of a 1% increase on targeted features for targeted segments. The matrix presents user segments as rows; the main KPIs as columns (groupable, hideable). A macro layer calculates the impact on revenue of a 1% lift of any chosen KPI, for each segment.
This makes it visible which segment would bring the most revenue from a feature — orienting not only which project to do first, but also which user segment we should keep most in mind.
Placeholder
Insert the live matrix here once the model is wired to the data warehouse.