How to attribute a sale to an AI recommendation
Knowing AI drives sales for your brand doesn't tell you which product benefits: here's how to tie a specific sale to a specific AI recommendation
2 min read
A brand can know ChatGPT cites it in general, and even measure AI-influenced revenue at the store level. But knowing which specific product got bought because it was recommended by AI is a different question, and a much more useful one for deciding where to invest next.
Without that granularity, you're optimizing blind: fixing product pages somewhat at random, hoping it counts, without knowing which ones actually drive sales once recommended. FIX budget gets spread thin instead of concentrated on the products that matter most. A limited FIX budget should go to the products that convert once recommended, not to whichever ones happen to get the most internal attention.
Why brand-level attribution isn't enough
AI-influenced revenue measured at the whole-store level confirms the channel works. But it says nothing about how it's distributed across your products. Two brands with the same overall number can have opposite realities: one where a single bestseller carries the whole effect, another where it's spread across ten SKUs. Without product-level detail, both teams would make the same decisions, when they should be prioritizing very differently.
How to tie a specific sale to an AI recommendation
- Spot the Shopify order whose origin carries the chatgpt.com referral.
- Identify the product or products purchased in that specific order.
- Connect that product to its AI visibility state at the time of the order: was it cited, recommended, and in what context.
- Consolidate these attributions product by product to surface your real winners, not just your usual bestsellers.
- Use that ranking to prioritize your next FIX efforts on the highest-potential products, rather than the whole catalog.
A sale that follows a click from ChatGPT is a strong signal, but a single instance stays fragile. It's the repetition, product after product and month after month, that turns the link into a reliable pattern rather than a coincidence.
That's the granularity Tadow is built for: connecting a real sale, placed in your Shopify store, to the AI visibility state of the product involved, closing the MAP, FIX, and PROVE loop down to the product level, in euros.