"We looked at a number of competitors in the space, but ultimately chose ThoughtMetric because of its easy-to-understand interface and the support offered during and after implementation."
"With ThoughtMetric, we were able to refine our analytics and provide verifiable proof of the revenue we were driving in a previously underperforming area of the business."
How to Use ChatGPT and Claude to Find Wasted Ad Spend in ThoughtMetric
Alex Fusco
July 09, 2026
Last Updated:
To find wasted ad spend using AI, connect ThoughtMetric to ChatGPT or Claude and ask it to surface campaigns with meaningful spend and little attributed revenue. Connecting takes a couple of minutes using the setup instructions in ThoughtMetric. From there, ask one question in plain language and get a ranked list of the campaigns losing money.
This is a great tactic for brands and agencies running dozens or hundreds of campaigns at once, where the underperformers are easy to miss.
Why wasted spend is hard to catch at scale
When you run a handful of campaigns, you can glance at the dashboard and spot the low performers. When you run hundreds across Meta, Google, and TikTok, and multiple stores, they hide in the volume.
AI connectors make this analysis easier
ThoughtMetric connects to ChatGPT or Claude throughAI connectors, letting you query your attribution data in natural language. Once connected, there are no exports and no pivot tables. You ask a question and get an answer.
Here are a few examples of what you can ask:
"Show me every campaign that spent more than $500 in the last 30 days but drove less than $500 in attributed revenue."
"Rank all my active campaigns by attributed ROAS for the last 30 days, lowest first. Flag anything under a 1x return."
"Which campaigns have steady spend but bring in very few new customers?"
Try it with your own data
The best way to see the value is with your own store's numbers. Within a few questions, you will know which campaigns to pause, which to rework, and which to fund.
