ChatGPT Ads traffic: send it to your Search landing page or a page built for the chat?

Where should ChatGPT Ads clicks land? The usual answer is to reuse the page your Search campaigns already send people to, since it's tested and converts. I'm not convinced that fits, because someone who got a recommendation inside a chat often arrives with a different question than a Google searcher. The chat has already done the comparison for them, so a long generic homepage may lose them. I'd lean toward a short page that repeats the exact product or offer the chat mentioned, with the same tracking setup so the numbers stay comparable. I'm not sure the lift is real or just a guess on my part. Has anyone split the two pages with enough volume to say?

Nobody I know of has run a split test with enough volume to call it, so I'd be cautious about any number someone quotes. It depends on how specific the chat's recommendation was. If the assistant named a product and a price, a matching page probably wins. If it just said "a good CRM for small teams," the generic page may still do the job.

Agreed, nobody has the volume to call it. The chat's wording often shows up in the search query or the referrer, so you can usually tell what kind of recommendation sent people. If they arrive with a product name in the query, the matching page is the safer bet.

Nova, I'm not sure the query part holds. With Google search the person types their own words, so the query tells you what they wanted. A chat ad click doesn't work that way. The product name the assistant mentioned may never show up in your analytics, and the referrer is often just the chat domain or blank, so you can't really read the recommendation from the traffic. Give chat-led clicks their own UTM value and, if you can, a separate landing URL, then compare both groups on the same conversion event. That will tell you a lot more than guessing from the query.

Agreed, Sasha, the query tells you very little for chat clicks. One thing to watch once you build that separate URL: assistants often quote prices from older pages, review sites or cached listings. If your dedicated page shows a different plan or a newer price than the chat mentioned, visitors bounce fast, and it looks like a weak page when it's really a mismatch. Keep the offer copy identical to your pricing page and recheck it after every price change. It's the dull part, but I'd expect a lot of the early drop-off to come from there.

It really comes down to how much chat traffic you get. With only a few dozen clicks a month, a separate URL just splits a small sample into two groups that are each too small to read. You'd spend weeks keeping a second page in sync with pricing and still have no idea whether it converts better.

At that volume I'd keep the Search landing page, give chat clicks their own UTM, and compare conversion rate against the Search group on the same event. A dedicated page only becomes a fair test once both groups have real conversions. Until then it's mostly one more page that can drift out of date, which is the maintenance problem you described.