Thread by @tonythill

I’ve been using ChatGPT’s new Code Interpreter to increase the CTR of my highest-ranking but lowest-clicked pages in Google.

Once it knows which title tags perform best for my site, I find it suggests much better ones🤌

Here’s how I do it in 7 steps 👇

Step 1: Get Data From GSC

In Google Search Console, export URL SERP performance data (CTR, impressions & clicks).

Step 2: Get Title Tags

Copy all the URLs from the sheet and paste them into this tool to extract all the title tags: exadium.com/tools/metadata…

Step 3: Combine Title Tags with GSC Data

Copy the title tags into the GSC spreadsheet in a new column.

(The title tags are listed in the same order as their URL 👍)

Step 4: Upload csv to Code Interpreter

Open up ChatGPT GPT-4 and enable Code Interpreter and upload the updated GSC spreadsheet.

Step 5: Identify Top-Performing URLs

Prompt: Identify the characteristics of the top 10% of URLs with the highest number of clicks and CTR. Tokenize the titles. Then look at the topics, keywords, length, structure, and other elements that could be contributing to their success

(Let it run its analysis)

Step 6: Identify Lowest-Performing URLs

Prompt: Identify the articles with the lowest CTR with at least 500 impressions and the position value between 1 and 4. My goal is to improve the title tag for these articles that rank high in Google but have a low click-through rate.

(Let it run its analysis)

Step 7: Suggest New Titles

Prompt: Give an idea for an improved title tag for each low-performing URL you listed based upon the analysis of the highest-performing article title tags.

(Now go update your title tags)

Tools like ChatGPT and Claude are making it way easier to analyze data in order to figure out what to double down on.

Title tags is just me dipping my toes in the Code Interpreter water...

Give this one a shot and let me know what ways your using Code Interpreter for SEO 🙌