How to run a UGC audit: coverage, rights, freshness, and what it is worth

An audit answers a question the dashboard does not: is the UGC programme doing its job across the catalogue, or only on the twelve products someone remembered to tag? Most brands run UGC as a series of moments. A campaign, a widget on the homepage, a gallery on the hero product. The audit turns that into a map of the whole store, with four measurements that can be re-run next quarter and compared. It is the same discipline as a stock count. Boring, periodic, and the only way to find out what you actually have.

The four measurements are coverage, rights, freshness and value. The first three are inventory questions: how much, is it usable, is it current. The fourth is the one finance asks. Run them in that order, because value means nothing if half the content counted is not cleared to show.

Coverage: how much of the catalogue has proof

Coverage is the share of products with at least one piece of customer content that is cleared and live. The naive version counts SKUs. The useful version weights by traffic, because a store where 15% of products carry UGC can be a store where most sessions see it, or a store where almost none do. Pull the product list, join it to sessions for the last 90 days, mark each product covered or not, and report both the SKU share and the traffic share. The gap between the two is the first finding of the audit.

Then cut it by category and price band. Coverage clusters where the marketing team already looked: the hero range, the launch, whatever was on sale. The long tail, which is often where margin lives, tends to have nothing. That list, sorted by traffic descending, is the collection brief for the next quarter. If the answer to "where do we get content for those" is a blank, the own-versus-rent library piece is the argument for building the pipeline rather than licensing stock every season.

Rights: cleared, dated, and not expired

Every piece counted as covered needs a rights grant that exists, names its source, and has not expired. Count the rest as uncleared, including the ones with a grant but no date, because an undated grant is a grant whose expiry you cannot defend. The reason is not only legal exposure. A creator who revokes, or a grant that lapses, pulls the content from every surface at once, and a coverage number that includes it is a number that can drop overnight.

The output here is three counts per product: cleared and current, cleared but expiring within 90 days, and uncleared or unknown. The middle column is the operational one. It is the renewal list. The rights-expiry audit covers the mechanics of finding these past a few dozen agreements, and the short version is that a shared inbox stops working long before the volume feels large.

Freshness: the age of the newest post

Freshness is the age of the newest cleared post per product. It is the measurement most audits skip and the one shoppers notice. A gallery whose most recent post is two seasons old tells the visitor the product stopped being bought, which is the opposite of the message the gallery exists to send. Set a threshold that fits your cycle, 90 days for fashion and beauty, longer for furniture, and report the share of covered products whose newest post is older than it.

Freshness also feeds the collection brief. A product with twenty posts and nothing new in a year needs a nudge (a post-purchase ask, a hashtag, a QR on the packaging) more than a product with three posts from last week. Sorting the covered list by newest-post age puts the stale-but-important products at the top. The competition runbook is the fastest way to refill a stale product; a monthly winner keeps the newest-post date moving.

Value: engaged visitors versus the store average

The value question has a clean version and a flattering version, and the audit should run the clean one. The flattering version compares pages with UGC against pages without, which mostly measures the fact that you put UGC on the pages that were already selling. The clean version compares, on the same pages and in the same window, the conversion of visitors who engaged with the UGC (opened a tile, played a video, tapped a product) against the store average for those pages. Engagement is a choice the visitor made, so this still carries selection bias. But it is the honest starting point, and it is the number a holdout test then confirms or corrects.

Keep the revenue lines apart. Attributed revenue is orders reconciled against a UGC click through the store's order webhook. Influenced cart value is the value of add-to-cart events that followed an engagement, whether or not they closed. They answer different questions and they are not additive, so a report that sums them is a report nobody should sign. Engagement metrics belong in a third column, because they exist for a brand with no commerce integration at all.

"Value delivered", the phrase that ends up on the slide, is then a sentence rather than a number: this share of traffic saw customer content, those visitors converted at this rate against this baseline, and this much revenue reconciled to it. Published vendor datasets show the ceiling.

Treat those as the top of the range. A lift measured on a catalogue with 15% coverage and stale galleries will not look like one measured across two hundred thousand stores that chose to install a reviews app. The audit exists to say which of the two you are.

Where you are, and what to do next

Running the first pass in an afternoon

Two free routes shorten the first pass. The UGC audit scanner does the outside-in check on a public store: which sources are visible, whether galleries are present and how heavy they are, whether rights and schema signals are exposed. The 30-minute audit is the inside-out version with a person on the call: widget weight, source coverage, rights state, layout fit, the attribution gap. Neither replaces the four measurements above. Both tell you which one to start with.