Amazon Attribution: proving social UGC drives marketplace sales

What Attribution measures, and the window it measures in

The reporting model is last-touch with a 14-day window: a shopper who clicks your tagged link and buys within a fortnight is credited to that tag, whatever else they saw along the way. It is a blunt model, and arguing with it misses the point. Attribution's job is not to settle multi-touch philosophy; it is to replace "we think social drives Amazon sales" with a number your finance team can look at. Before Attribution, off-Amazon campaigns pointing at listings vanished into a measurement hole: you could watch the click leave Instagram and watch the rank move weeks later, with nothing in between. The console fills the in-between.

The funnel metrics matter as much as the purchase count. Detail page views tell you a creator's audience was curious; add-to-carts tell you the listing survived their scrutiny; the gap between the two tells you whether your problem is the traffic or the PDP it lands on. A campaign that produces heavy detail page views and thin purchases is not a failed campaign, it is a diagnosis, and often the prescription is better evidence on the listing itself: the A+ and Posts work covered on the Amazon syndication page.

Tagging UGC campaigns: granularity is the whole game

The default mistake is one tag per channel: a single "Instagram" tag on everything, producing a number nobody can act on. Attribution supports tags per campaign, per ad group and per creative, and UGC programmes should use that structure the way they already use UTMs: one tag per creator, per post, per placement. Tagged that way, the console starts answering the questions that decide budgets. Which creator's audience actually buys on Amazon rather than just engaging? Does the unboxing format outconvert the try-on? Do story clicks window-shop while feed clicks purchase? Aggregate tags cannot answer any of these; granular tags answer all of them by default.

Naming discipline is the unglamorous half of this. A tag called "sarah-tiktok-unboxing-aug" is a datapoint; a tag called "campaign-3-final-v2" is a future archaeology project. Decide the convention before the first tag ships.

The loop: measurement feeding the content engine

Attribution turns a UGC programme from a broadcast into a loop. Content goes out tagged; detail page views, add-to-carts and purchases come back per creator and per asset; the next round of briefs, whitelisting budgets and rights requests goes to the people and formats that demonstrably moved Amazon shoppers. And the winners earn a second life beyond the campaign: the clips proving themselves in Attribution reports are precisely the ones worth cutting into A+ modules, Posts and Sponsored Brands Video through the syndication pipeline, provided their rights scope covers those surfaces. The loop closes when the content that sells on Amazon is also the content shown on Amazon.

The Brand Referral Bonus: measurement that part-funds itself

There is a direct financial kicker for doing this properly. Amazon's Brand Referral Bonus program credits brands a bonus averaging 10% of qualifying sales driven by tagged off-Amazon traffic; the 10% average is Amazon's own published figure, and the actual rate varies by category. The mechanism runs on Attribution tags, so a brand tagging its UGC campaigns at proper granularity is simultaneously building its measurement dataset and clawing back a slice of referral fees. For a programme already paying creators, the bonus can offset a meaningful part of the cost of running the loop at all.

Where our numbers end and yours begin

A disclosure we make every time, because the difference matters. The conversion figure we quote for Idukki (a lift of roughly 20%, confirmed in VWO-run A/B tests) was measured on Idukki customer storefronts. It is an on-site number: shoppers seeing UGC galleries on brand websites and converting there. It is explicitly not an Amazon-measured number, and nobody currently has an Amazon-measured equivalent to sell you, because Amazon does not run your listings as an A/B test on your behalf.

That is precisely what Attribution is for. Tag your UGC campaigns, point them at your listings, and within a quarter you will hold something better than any vendor benchmark: your own Amazon-side number, on your own catalogue, at your own price points. The mechanism our on-site tests confirmed (shoppers trust evidence from people like them) has no reason to stop at Amazon's front door. The size of the effect on your listings is an empirical question, and the console that answers it is free.