Measurement ·
Before you test ChatGPT ads, what should you set up so you can judge them fairly?
An afternoon ChatGPTGoogle
Short answer
Tag paid ChatGPT clicks as their own campaign, then compare revenue per visit with your own paid Google Search over the same weeks. Turn that ratio into a break-even price per click before you agree to a bid.
A working paper by Kaiser and Schulze, covered by Search Engine Land on November 2, found that in US traffic a paid ChatGPT visit brought in $0.89 against $2.45 for a paid Google Search visit across the advertisers it tracked. That is one study of mostly general online stores, not a verdict on your brand, and the authors note the paid comparison is descriptive. It does give you a method. If you track the ads separately and compare them with your own search results, you will know within a few weeks whether the price you pay per click makes sense. Without separate tagging, paid ChatGPT visits can blur into organic ChatGPT referrals, which the same paper found kept converting at their old rate.
Do this
- Give every ChatGPT ad its own UTM campaign and source, and check in GA4 that paid visits land in a separate row from organic chatgpt.com referrals.
- Run the test for at least four weeks, alongside your regular paid Google Search campaigns, so both cover the same weeks and promotions.
- Divide revenue by sessions for each channel. Then divide the ChatGPT figure by the Google Search figure to get a ratio.
- Multiply that ratio by your average Google Search cost per click. The result is the most you can pay per ChatGPT click for the same return, before margin differences.
- Compare that number with what you actually pay per ChatGPT click. If you are paying several times your break-even, cut the test or narrow it to the products that did convert.
Effort is our estimate for a small brand team. Assistant tags come from the assistants this tip names. See also AI shopping assistants compared.