Do you think tools like OpenAI’s Codex or Anthropic’s Claude Code are developer tools, designed only for writing software? This is not the case. A recent project at SmarterX shows how these tools can be repurposed for one of the most common (and tedious) marketing tasks: making sense of messy data.
The problem
The goal was to understand how a specific piece of content relates to revenue. The data was there, but the answer wasn’t organized into a single field. It was buried inside a sprawling export plot: 144,000 rows and 1,000 columns. The file was so large that simply opening it in a spreadsheet program would crash the computer trying to load it.
approach
Instead of manually manipulating pivot tables or dropping a static file into a chatbot for a one-time summary, the entire data set went to Codex, where it is used by an analyst sitting next to the work. Using a fully anonymized export process, Codex is tasked with:
- Examine the data and identify fields that appear revenue-related versus attribution-related
- Flag fields that were too annoying or duplicate to be trusted at face value
- Conduct safety checks on smaller groups to validate their approach before scaling up
- Narrow down 1,000 columns to a focused, relevant group
The result was a clear path to revenue attribution modeling, without writing a single formula or manually sorting columns.
Why this is important for marketers
The value here is not that Codex wrote the code. The tool can be set a goal, “find what correlates with revenue,” rather than a list of manual steps, and the agent performs its own multi-step investigation. She determined her next steps, corrected her errors, and continued working until the analysis was complete.
This is a completely different way of working than typing questions into a chat program one by one and waiting for individual responses. It’s closer to delegating a project to a capable analyst than to search engine prodding.
Marketers don’t need to be developers to take advantage of this. Any team working on a messy CRM export, a tangled campaign performance report, or an attribution data set that no one has time to dig through can use proxy tools like Codex or Claude Code to work through that investigative process. The starting point is not a perfectly clean spreadsheet or precise formula; It’s a clear goal and a willingness to let the agent know the path.
To hear the full use case explained and more about what’s happening with AI, check this out Episode 222 of the Artificial Intelligence program Podcast.
(function(d, s, id) {
var js, fjs = d.getElementsByTagName(s)(0);
if (d.getElementById(id)) return;
js = d.createElement(s); js.id = id;
js.src = “//connect.facebook.net/en_GB/sdk.js#xfbml=1&version=v3.0”;
fjs.parentNode.insertBefore(js, fjs);
}(document, ‘script’, ‘facebook-jssdk’));
