Glossary
What is AI-Attributed Conversion?
Definition
AI-attributed conversion is a sale or signup credited to a visit that began with an AI assistant naming your brand. Unlike a standard last-click conversion, which has a referrer to trace, an AI-attributed one is usually inferred from survey answers, landing behavior or a gap in direct traffic. The method you choose decides the number.
At a glance
- CategoriesMeasurement
- Related fieldsAnalytics, Commerce
- DifficultyAdvanced
At a glance
- Most AI-sourced orders arrive with no referrer, so the assistant that caused them leaves no trace in analytics.
- Three methods dominate: a self-reported question at checkout, referrer matching, and modelled lift against a baseline.
- Self-reported attribution costs a dropdown on the order form and survives a device switch.
- No standard definition exists, so two vendors reporting on the same store will produce different figures and both will be defensible.
The chain from an answer to a paid order
A shopper asks Perplexity which brands make a quiet dishwasher. The answer names four, yours included, with a link. That is the only moment the assistant is involved.
What happens next is where attribution breaks. Some shoppers click, and you get a referrer you can read. Many more read the name, close the tab, and come back two days later by typing your brand into a browser bar. A few switch device in between.
Only the first group appears as AI search traffic. The rest arrive as direct or branded search, and they buy readily, because they came back on purpose. The orders most clearly caused by the assistant are the ones least likely to be credited to it.
Undercounting makes your cheapest channel look like your worst
Budget follows measured return. A channel that produces orders it cannot claim gets defunded by a spreadsheet, not by a decision.
The shape is familiar. A brand runs six months of work on AI visibility, sees a flat line in the channel report, and quietly stops. Direct traffic and branded search both rose over the same period, and nobody connected the two.
The reverse error costs as much. Credit every direct order in a good month to AI and you will fund the wrong thing on a number you made up. Pick a method, write it down, hold it still long enough for the trend to mean something.
Get a defensible number this month
No platform required. What it does need is a decision about method and the discipline to keep it.
- Add one question to your order confirmation. “How did you first hear about us?” with options naming ChatGPT, Perplexity, Gemini and Copilot, plus free text.
- Tag the referrals you can see. Give the assistant domains their own channel grouping instead of letting them fall into referral or direct.
- Record a direct-traffic baseline before you start any GEO work. The comparison is worthless without a before.
- Cross-check against the prompts you appear in. If you became a cited source for your top buying prompts in April, look for the direct-traffic move across April and May.
- Report the method beside the number. A conversion figure with no stated method invites the first person who dislikes it to dismiss it.
Three ways to credit an AI-sourced order
| Method | What it catches | Where it fails |
|---|---|---|
| Self-reported at checkout | Buyers who never clicked, across devices | Only fires on buyers, and recall is imperfect |
| Referrer matching | Clean click-throughs from assistant domains | Missing referrers, in-app browsers, copied links |
| Modelled lift on direct | Aggregate effect with no buyer input | Confounded by every other campaign you run |
| Prompt-appearance timing | Directional link between visibility and revenue | Correlation only, never a per-order credit |
Frequently asked questions
-
Is AI-attributed conversion the same as last-click attribution?
No. Last-click attribution credits whichever touchpoint immediately preceded the order, which for an AI-influenced buyer is usually direct or branded search. AI-attributed conversion tries to credit the assistant that put you on the list in the first place. The two will disagree, and the last-click number will almost always be the smaller one.
-
How do I start measuring this without buying anything?
Start with the checkout question. A dropdown on your order confirmation naming the main assistants gives you first-party data within a week and captures buyers who never clicked a link. Add the assistant domains as a named analytics channel on the same day. Both changes are under an hour of work.
-
Is this worth doing for a small brand?
Not yet, if your order volume is low. Self-reported attribution needs enough orders for the percentages to stop jumping around, and below a few hundred orders a month the noise swamps the signal. Track appearance in assistant answers instead, which works at any volume, and add conversion attribution when the sample supports it.
-
Can I trust a vendor's AI conversion number?
Only once you know how it was built. Ask which method sits underneath it, what happens to a buyer who switched devices, and whether the baseline predates your campaign. A vendor who cannot answer those in plain language is selling a modelled estimate as an observation.