From ad results to learnings
Today. Lup reads your Meta Ads results: spend, attributed purchases and revenue, ROAS, CTR, CPC and CPM, by campaign and against the previous period.
Next. We’re building the reading of each Experiment: which variant looks promising, the likely reason, and how sure Lup is.
Get startedFictional brand. The Experiment screen is a preview.
Full frameImage · Feed 4:5$420.0014$532.00 revenue1.27×Tight cropImage · Feed 4:5$420.0021$798.00 revenue1.90×Each variant, day by day
Lup will compare variants on the metric the Experiment declared.
A likely reason, with its limits
Each learning will keep its evidence, a likely reason and how sure Lup is.
Preview of what we’re building, with a fictional brand.
How an Experiment can end
Promising result
One variant led on the chosen metric for most of the run. Lup will name it and give the likely reason.
Too early
Too little spend or too few purchases to compare. Lup will say so instead of picking a variant.
No real difference
The variants performed alike. Lup will close it without a conclusion and record what was tested.
Preview of what we’re building, with a fictional brand.
Today and in progress
Today
- Meta Ads results in Analytics, read-only.
- Spend, attributed revenue, ROAS, CTR, CPC and CPM by campaign.
- Comparison with the previous period.
Building
- Reading each Experiment’s variants.
- Learnings with evidence, a likely reason and limits.
- Next variations that start from a learning.