I’ve seen technology upend how we create and share content more than once in my career. Starting out as a print journalist, I watched the internet and social media completely reshape media companies. Years later, my health blog took a massive hit after Google’s Helpful Content Update.
Currently, we’re seeing AI change how people do research online. If you’re creating content online, you’ll know AI tools can help with a whole range of tasks: keyword research, content briefs and topic discovery. But what makes the difference is how you use AI insights alongside real user data.
Here are 4 steps to combine AI with Google Analytics 4 (GA4) to create a content strategy that drives real traffic and engagement.
Step 1: Check Your AI Content Traffic Sources in GA4
Understanding where your visitors come from is essential. GA4’s interface can be tricky, so here’s a quick way to see which AI channels deliver the most traffic to your content:
- Open Google Analytics 4
- Go to Reports > Acquisition > Traffic acquisition
- Select Session source / medium from the dropdown above the table
From here, you’ll see exactly which LLMs are sending traffic to your site.
Step 2: Use Perplexity.ai to Analyse Content Sources
Perplexity’s free AI search tool is great for quickly checking what sources feed content. Because Perplexity shows all of it’s sources, even on the free plan, you can check which sites are being pulled into LLM results. Review the results and see what patterns emerge across featured articles.
This insight lets you tailor your content to compete but with your unique voice and perspective.
Step 3: Use LLMs to Find Link Opportunities and Training Data Insights
Ask AI tools about your brand and the websites they learn from. This approach will help you discover new link opportunities and figure out how to get noticed by LLMs.
Marie Haynes, in this Clearscope SEO panel, suggests that if you are working with a brand and want it to be recognised by Large Language Models (LLMs), such as ChatGPT or Gemini, you should start by asking the LLM questions about the brand.
Specifically, she recommends asking:
• What brands can you recommend for this (product/service category)?
• If your company is not mentioned, you should then follow up by asking: “Why didn’t you mention my company? What can I do so that you’re more likely to mention me?”
In addition, Rand Fishkin recommends directly asking an LLM, such as ChatGPT or your LLM of choice, about the odds of specific websites being used in their training data.
For example, he demonstrates asking: “On a scale of zero, not likely at all, to a hundred, almost certain, how likely are the following websites to be used in the training data of large language models like Gemini and ChatGPT and Perplexity”. After providing a list of sites, the LLM will give an estimate.
This approach offers a way to gain insight into potential sources used in an LLM’s training data. You could, for example, approach a few of the sites mentioned and try to get your site featured in the roundup.
Step 4: Refresh Your Content Regularly to Combat Recency Bias
Google favors fresh content. Even small updates signal your page is current and relevant.
SEO expert Steve Toth explained in an interview that AI tends to prefer more recent content.
Toth pointed to Perplexity’s leaked system prompts containing explicit wording about preferring more recent results. If your company is one of the few with a very recent change, you’ll up your chances of being included in LLM responses.
By checking your traffic sources in GA4, using AI tools like Perplexity and ChatGPT for research and feedback, and keeping your content fresh and relevant, you can build a content strategy that truly drives growth.
