Measuring the return on UGC is harder than dashboards suggest, because the same video works in several places: in ads, on the product page, in emails. Attribution assigns it to one channel and is wrong almost every time.
The three numbers that actually count
- Cost per acquisition of the campaign using the video, compared with the same campaign before the video. An imperfect comparison but a readable one.
- Product page conversion rate with and without video, measured over at least two weeks per side.
- Content cost divided by acquisitions during the period it ran. If a video costing a few hundred euros accompanied hundreds of orders, the return question stops being interesting.
Why attribution lies
Someone sees the video on TikTok, searches the brand on Google two days later and buys. The dashboard assigns the sale to search. That is not a technical error, it is the limit of the model. This is why a before and after comparison across a whole period is more reliable than per channel attribution.
The simplest test that works
Keep one variable moving. Two weeks with the old creative, two weeks with the new one, same budget, same offer, same audience. The difference in cost per acquisition is your data, and it is sturdier than any model.
A common mistake in Italy
Judging a video by its likes. Effective ad content often has mediocre comments and excellent sales, because it is not made to be liked but to make people buy. The two metrics are not correlated.
The cost nobody counts
Internal time. Briefs, approvals, revisions and uploading cost more than the video itself in many Italian companies. Cutting approval rounds is often the strongest return lever available.
Keep reading
- UGC for Supplements and Fitness Brands in Italy
- UGC for Travel and Hospitality in Italy
- Using Creator Video in Email and Retention in Italy
- UGC Creators in Palermo
- The complete UGC guide for Italy
Find creators and measure the result →
Frequently asked questions
Which metric should I read first?
Cost per acquisition, before and after.
Do likes matter?
No, they do not correlate with sales.
How long is a valid test?
Two weeks per variant.
How do I measure the product page effect?
Compare conversion rate with and without video.
