As a follow-on to yesterday’s post about AI search strategy, I wanted to provide a practical example of what this looks like.

There is a lot of attention being given to visibility in AI search engines such as ChatGPT, Perplexity.ai, and others. This is going to bring questions from stakeholders outside of traditional SEO teams as they begin to understand the importance attracting traffic via these new channels. We view this as a great opportunity for SEO to break out of its traditional silo and act as a key advisor for AI strategy at many types of businesses.

This is easier demonstrated with concrete examples.

Up until the launch of ChatGPT and its integration of the Bing search index in its result (particularly in its search feature), in our industry Bing was a search engine that did not have a lot of demand for tracking.

Nevertheless, we’ve always tracked it at a basic level and are adding a ton of additional support and features for it now, precisely because of the reasons mentioned in this post.

So, I added Bing as a search engine to one of the tracking instances we have for the demandsphere.com domain, and pulled up the following ranking report:

So, as of the date of this post, we rank position 1 for the term “large scale rank tracking” on Bing.

We can test the theory that ranking well in Bing translates to strong visibility in ChatGPT:

As you can see, our page is referenced extensively in the ChatGPT results for the prompt “I’m interested in large scale rank tracking.”

Questions of search volume, demand, etc. are also important but still orthogonal to the point we’re making to today about the mechanics of how this works.

The point is this: if you want to appear in ChatGPT Search, you have to appear in Bing for the topics you care about.

One of the interesting things to note about search in ChatGPT is that results are not updated daily.

My post from yesterday serves as a good example of this.

We’re currently ranking #1 on Bing for “ai search strategy” (although I expect this to drop), yet if you prompt ChatGPT search for “tell me about ai search strategy in 2025” I get the following results:

This is the same results that I got yesterday as well, so there is some caching involved in search results coming from ChatGPT. I verified this by testing several variations on this prompt and got the same results every time.

I’d guess that these results are cached for a few days at a time, but will keep an eye on this to be sure. This means that weekly monitoring on these type of results is likely sufficient for now. I expect this to also change over the long term.

We are still early in the lifecyle of AI search engines and monitoring is still very immature.

The few tools that I know that can do it are relatively new and focused primarily on these new search experiences vs. the traditional search engines. Over time, as more tools in our industry add in monitoring capability for ChatGPT, Perplexity, and others, the point solutions will lose a bit of their novelty and will need to find additional ways to differentiate.

In the meantime, I still believe that your data strategy is best focused on Google and Bing, because of the downstream impact they have on the LLM-based search engines.

There are several additional advantages of tracking and optimizing directly for Google and Bing:

Scale and volume

ChatGPT and others make it very difficult to collect data from their web platforms. And you can’t get the same results via their API, even if you RAG in Bing results (although you can get kind of close).

For this reason, there exists a data collection bottleneck which constrains the amount of volume you can reliably monitor directly. For large, enterprise sites, this is a problem. The good news is that with tools like ours and others, large-scale SERP monitoring for Google and Bing is a much easier prospect.

Real-time / daily updates

Given the above discussion about caching of results in ChatGPT, you’re not going to get daily updates on all of your prompts, regardless of how you monitor. But if you’re monitoring Google and Bing daily, you’ve just established a predictive / leading indicator of how results in ChatGPT are likely to evolve.

Leverage existing SEO knowledge and operations

The advantage of this approach is also further bolstered by the fact that SEO teams already know how to do all of this. They have the data, processes, and team members in place to hit two birds with one stone.

Things will continue to evolve and it won’t always be this clean and connected but why not take advantage of the situation as it is today?

A final topic worth considering in this context is Perplexity.ai, since they are one of the few non-Google search engines that do not appear to be using Bing (at least primarily) for its search engine functionality.

As I mentioned yesterday, I believe that they are relying heavily on Google.

Fortunately, for the sake of this example, we perform decently well for the same keyword “large scale rank tracking”:

We can see this by checking Google directly as well:

To my-not-surprise, we are showing up nicely in Perplexity.ai for the same prompt “I’m interested in large scale rank tracking”:

When investigating the 6 sources on the right-hand side, these sources appear prominently in the Google SERP for “large scale rank tracking.”

So the basic idea holds and is worth experimenting on your own with this further.

I’d love to hear from you if you get different results or find anything else new and interesting.

All of this further reinforces what our data study and white paper from last month revealed when evaluating AI Overviews on Google. Google is obviously relying on their own search index as a key means for filtering sources for inclusion in AIO citation links.