ChatGPT Ad Targeting Criticized for Low Conversion Rates and Poor Traffic Quality

Experimenting with ChatGPT Ads

An advertiser conducted an experiment running ads on ChatGPT for event seating assignment software from August 5 to August 28, 2026. The advertiser bid between £0.07 and £0.15 per click, significantly lower than ChatGPT's recommended £2 per click, yet still acquired traffic.

Low Conversion Rates Observed

The experiment resulted in a 0.46% conversion rate, which the advertiser deemed insufficient to generate profit given the product's lifetime customer value. This conversion rate was considerably lower than traffic from other sources during the same period, suggesting an issue specific to ChatGPT ad traffic.

Investigating Potential Causes

The advertiser systematically investigated several potential causes for the low conversion rate. Discrepancies between OpenAI's reported clicks and Google Analytics were ruled out, as were issues with the ad's context hint or the ad creatives themselves, which were similar to successful Google Adwords campaigns. Initial concerns about bot traffic were addressed by consulting digital ad fraud expert Dr. Augustine Fou, whose analysis indicated that while some bot activity was present (30% showing red flags), it was not extensive enough to explain the poor conversion rate.

Conclusion: Poor Targeting

With other factors ruled out, the advertiser concluded that poor traffic targeting was the primary cause of the low conversion rates. Fouanalytics data supported this, showing that 88% of traffic moved their mouse but only 10% clicked anything on the landing page, indicating a lack of genuine interest from the audience receiving the ads.