Measuring on-site UGC: views, clicks and attributed sales
How to tell whether the creator content on your site is earning its place. The metrics that matter at each step, from wall impressions to product clicks to attributed orders, and the traps that make the numbers look better or worse than they are.
Putting creator content on your website is the easy part. Proving that it earns its place is harder. Many teams install a UGC gallery, see a big impressions number, and stop there. Impressions say the block loaded. They do not say anyone cared, and they say nothing about sales.
This post sets out a simple measurement model for on-site UGC, the metrics at each stage, and the most common ways the numbers mislead.
Measure it as a funnel
On-site UGC does one job: it moves a visitor from "interested" to "on the product page" to "bought". Measure each step.
| Stage | Question | Metrics |
|---|---|---|
| Seen | Did visitors actually see the content? | Sessions with the wall, impressions |
| Engaged | Did they interact with it? | Card clicks (opening a post), filter use |
| Sent to product | Did the content send them to something they can buy? | Shop or product clicks, product click-through rate |
| Bought | Did those visits turn into orders? | Attributed orders and revenue, net of refunds |
Each stage tells you something different, and a weak stage points to a different fix.
- Low impressions per session: the wall is placed below the fold, or on pages few people reach.
- Many impressions, few card clicks: the content is not compelling, or the first tiles are weak. Pin your best post first.
- Many card clicks, few product clicks: posts are not tagged, or the tags are not visible enough.
- Many product clicks, few orders: the problem is likely the product page, price or availability, not the UGC.
The leading indicator: product click-through
If you track one number week to week, make it product click-through from the wall: product clicks divided by sessions that saw the wall. It is close enough to revenue to matter and frequent enough to react to.
Two diagnostics help here:
- Popular posts with no product tagged. A post that gets lots of opens but has nothing to click through to is a lost sale. Tag it.
- Products that get clicks but no sales. These are worth a look on the product page itself.
The result: attributed revenue
Clicks are a proxy. Revenue is the result. To connect them you need two things: a way to recognise a visitor who came through the wall or a creator link, and a feed of orders from your store, including refunds.
First-party attribution does this without third-party cookies. A small script on your storefront records the visit; your store's order webhooks report completed orders and refunds; the two are joined. Our creator attribution guide and the post on attribution after the cookie cover the mechanics.
Read attribution confidence, not just totals
Not all attributed sales are equally certain. A promo code redeemed at checkout is strong evidence. A purchase minutes after a click in the same session is good evidence. A purchase weeks after an earlier click is weaker. Report the tiers separately so you know how much of the total rests on strong evidence. The same tiers drive the maths in the full creator ROI measurement guide.
Traps that distort the numbers
Ad blockers and privacy settings
On-site analytics run in the visitor's browser. Ad blockers, strict privacy modes and some content security policies stop the events from reaching any analytics tool. Expect your UGC numbers to read lower than reality, and do not panic if one week reads zero while the wall is clearly being served: check the wall on a clean browser first.
Impressions without attention
A wall low on a long page will log impressions from people scrolling past. Look at card clicks and product clicks per session, not raw impressions.
Correlation is not lift
A product page with UGC may sell better simply because you put UGC on your best-selling products. To claim that the content caused extra sales, you need a comparison:
- A/B test the block. Show the wall to half of visitors on a page and not to the other half, and compare conversion.
- Use a holdout. Keep a set of comparable product pages without UGC for a period and compare.
- Compare like with like. Same products, same period, same traffic sources.
The glossary entry on holdout tests explains the method. Without a comparison, report what the wall did (clicks, attributed orders), not what it caused.
Refunds
Report revenue net of refunds. Fashion in particular has high return rates, and a wall that drives impulse buys that come back is not a win.
What to report, and how often
A short monthly view is enough for most teams:
- Sessions with the wall, and product click-through rate
- Top posts by product clicks, and popular posts with nothing tagged
- Attributed orders and revenue, net of refunds, split by confidence
- Changes made that month (new posts, re-tagging, placement) and their effect
Tie it back to creators. If the content that drives the most product clicks came from a particular creator, that is a signal for your next campaign.
How cyrqle measures walls
Each content wall in cyrqle has an Analytics step with four views:
- Composition: what the wall is showing right now.
- Engagement: sessions, impressions, card clicks and outbound clicks.
- Products: shop clicks and product click-through, plus a list of popular posts with no product tagged.
- Fan activation: a roll-up of posts that came from fan campaigns.
Wall clicks and store orders are joined in the campaign's attribution reporting, once your store is connected with the tracking snippet and order webhooks. Attributed sales are filed by confidence tier: promo code redeemed, purchase in the same session as a creator click, or a later purchase within a 30-day window. Revenue is net of refunds.
Like any browser-side analytics, wall engagement is invisible when a visitor's ad blocker stops it, so treat it as a lower bound.
FAQ
What is a good click-through rate for a UGC wall?
It varies too much by placement, category and traffic to give a single benchmark. Track your own baseline for a few weeks, then measure changes against it.
Can I attribute a sale to a specific creator's post on my site?
Yes, when the visitor clicks a tagged product from that post and the order is matched back. The link between the post and the creator is what lets you reinvest in the creators whose content sells.
Why does my UGC analytics tool show fewer sessions than Google Analytics?
Different tools count differently, and ad blockers affect them differently. Use one tool's numbers consistently for trends, and do not add numbers from different tools together.
How long should I run a test before judging a wall?
Long enough to cover normal weekly cycles and to collect a meaningful number of orders on both sides. For most stores that means several weeks, not days.
Related field notes.
A public post is not a licence. What brands in Spain, the EU and Latin America need before they reuse customer and creator content: copyright, image rights, GDPR, clear usage terms and a record you can prove.
How to turn rights-cleared creator and customer content into commerce surfaces on your own site: landing-page walls, social-proof grids and product-page carousels, with products tagged on every tile.
Returns-adjusted ROAS for fashion creator campaigns: gross vs net ROAS per creator with worked numbers, why commission waits for returns, and who changes rank.