# Lift study readout

*Template from the Vexpower MMM library: vexpower.com/mmm/conversion-lift-studies. Fill one in for every conversion lift or brand lift study, including the inconclusive ones.*

## Study basics
- **Platform and tool:** (Meta Conversion Lift / Meta Brand Lift / Google Ads Conversion Lift, users or geo / Google Ads Brand Lift / other)
- **Who designed and analysed it:** (platform, agency, in-house)
- **Campaigns and ad sets included:**
- **Dates:** start ___ end ___ (plus any post-period counted: ___ days)
- **Spend in the test group:** $___
- **Holdout / control share:** ___%
- **Randomization unit:** (user, household, geo)
- **Outcome measured:** (pixel/CAPI event, offline upload, survey question)

## What the result is a lift *of*
- **Comparison:** everyone assigned to test vs everyone assigned to control (intent to treat)
  OR exposed vs would-have-been-exposed (ghost ads / predicted exposure)
- **Share of the test group actually reached:** ___%
- **Reported lift is:** ITT / treatment-on-the-treated / not stated (ask)

## Results
| Metric | Point estimate | Interval (state the level, e.g. 80% or 90%) |
|---|---|---|
| Relative lift | | |
| Absolute lift | | |
| Incremental conversions | | |
| Cost per incremental conversion | | |
| iROAS (revenue) | | |
| iROAS (margin) | | |
| Brand lift: absolute lift (pts) | | |
| Brand lift: relative / headroom lift | | |
| Survey responses (test / control) | | |

- **Platform-attributed conversions, same campaigns and dates:** ___
- **Incrementality factor (incremental / attributed):** ___

## Limits to record
- Measures this platform only; conversions it can't see (offline, other devices, no consent) are missing.
- The platform designed, ran and graded the study.
- Anything that changed mid-test (budget, bidding, creative, audience):
- Seasonality or promotions in the window:

## Decision
- **What we decided:**
- **Will this feed the MMM?** (prior / calibration input / validation only)
- **What we'll test next and why:**