Most major search engines now have an AI search mode. But until recently, it was pretty much impossible to find out whether these AI search modes were using your writing for their answers.
That has changed now with Google providing Generative AI reports and Bing giving AI performance data. Don’t get too excited, though, as neither gives click counts as yet.
If you are publishing your writing online regularly, such as for your blog or website, it’s now at least possible to see which of your pages AI search loves to use. And by omission, which pages are mostly ignored.
Another interesting possibility is that you can now compare pages that get good results from standard search clicks with pages that are mentioned in AI search panels. And that comparison can reveal something you might not expect: the articles that attract readers from traditional search aren’t necessarily the same articles that AI search finds useful.
Writing and publishing are changing rapidly
Search engines with the traditional ten-blue-links search results are becoming less dominant. But a few like Qwant and Dogpile remain and have no AI functionality. DuckDuckGo is similar, but does offer AI as an opt-in.
Google and Bing have both made AI increasingly prominent in their search experience.
For writers and publishers, the effect can mean fewer clicks and less organic traffic.
Where these lost clicks went was a question with no answer until now. But if you use Google Search Console and Bing Webmaster Tools, you can get data about the pages AI accessed, and with Bing, also the queries that AI used.
I had my suspicions before, but now I can see clearly that Google AI search regularly uses many of my grammar and vocabulary articles for answers. These types of articles are prime fodder because AI can pull out a quick answer.
Examples from my Google data are mostly “What’s the difference between X and Y?” and “What is X?”
But my Bing results are totally different. Many of the most used terms for Copilot search come from queries like, “What can I use to do this?” or “What choices do I have to do this?”
Then there’s the difference in attribution in AI answers. Google AI Overviews have small links, image links, and sometimes an associated page widget.
However, Bing is far more generous with highly visible links in Copilot search results.
Until a few years ago, it was easy to write for your readers as well as search engines because a little SEO was all you needed to rank well. But now, how can you please standard search engines, AI search engines with differing priorities, and still please readers?
You can only write for one master
You can’t publish four versions of an article to try to please the machines and your readers.
Writing and publishing have always been reader-first, even if a few SEO tricks worked in the past without your readers knowing.
You might think you have to make a compromise now, but you don’t. The big difference is that you have access to enough data to decide what you write about.
My first thoughts when I started to get AI search data about my site were to think that AI was stealing my content and not giving me any clicks in return. Why wouldn’t I, when I can now see proof of around 200,000 AI citations of my articles per month on Google and Bing?
But digging a little deeper, I also discovered that around 75-80% of the pages on my site are being actively used in AI responses.
So if I wanted to avoid AI using my writing, I would have to shut down 80% of my site. On the other hand, it was a backhanded compliment that 80% of my site was so useful for AI answers.
What was a little surprising was when comparing my page view analytics with AI citations.
The top pages readers regularly visit are not the same as the top AI pages. Most are opinion, discussion, or how-to articles, along with a few listicles.
You can do your own analysis if you can access the data, but I’m sure you’ll discover a similar pattern.
Readers seem attracted to informative long-form writing, while AI might satisfy their need for short-form quick answers.
That doesn’t mean you should stop writing on topics or in a form that AI search likes.
Things can change rapidly, and you might earn more clicks in the future.
But until that change comes, deciding what you write about must be for your readers first.
Gaining readers is no longer only about search engines. Many other pathways can lead to all forms of your writing being discovered by readers.
Use your data to make better decisions
When you’re researching new ideas to write about, you need to take a little more care now. Gone are the days when you could simply hunt down a low-competition, high-volume keyword and hope for the best.
Today, reader intent matters much more. What do readers want to know, what do they want to do, and how can your writing help them?
Clearly, AI search is producing plenty of zero-click results for questions that only require a quick answer.
But when you look at your own analytics, do your readers still find these pages useful and worth visiting?
It was a lightbulb moment for me when I compared my AI citation data with my traffic data. Some pages that AI regularly uses for answers are also attracting readers from other sources.
Some of these pages often answered a question quickly, but also went into some depth about the choices, history, uses, and possibly pros and cons.
That makes AI data interesting and relevant, even if it doesn’t tell you how many people clicked through. It tells you what questions and interests users have that your writing answers.
It gives you another way to understand your site, your readers, and the different ways your writing can be discovered and used.
But don’t get carried away with the numbers. Statistics can tell you what might be happening, but data alone can’t tell you what you should write that will connect with readers. Only you know that answer.
The last thing you want to do is write for AI. But it’s certainly worth learning from what AI appreciates in your writing.
Author’s note on change
We don’t like change, do we?
But if the world resisted change, we would still be renting VHS videos and stacking 45s on a record player.
Today, technology changes at an astonishing rate, and much faster than we can immediately and readily learn to accept.
But that’s how it is. AI search is just one small part of how technology is changing what we do, how we react, and what we accept.
Like all writers, I originally saw AI as a threat. But it doesn’t matter now because it’s the reality that writers can’t avoid.
All you can do, as I have done, is accept that it’s a reality now and move on to finding ways to adapt and change.
Yes, I could wring my hands and pine for the past, but that wouldn’t help.
What I see in my AI data is actually a new positive. It shows that my writing is still useful because AI search values my writing and is citing it 200,000 times a month to satisfy users’ questions.
All I have to do is learn how to channel that interest back to my site. Unlike VHS tapes and 45s, my writing isn’t redundant at all.
It’s just in need of fresh and new ideas and ways for readers to discover it. And that’s my challenge now.
Summary
Despite a lot of negativity about AI search, it is users who are voting with their new search habits.
Increasingly, users search for an answer rather than a list of ten links they have to click through.
If you’re a writer, you’ve probably adapted your searches as other users have done. So the next step is for you to adjust your writing to suit the realities of the new search world.
The best way to do that is with useful data and analytics. And yes, the AI data you can get is limited in scope, and minus any click data.
However, I’m sure that will evolve and improve as AI search matures. For now, though, at least you now have some data showing you how AI is using your writing and publishing.
In the meantime, keep writing for your readers, and let AI do what it does, and cross your fingers and hope that it gives you a link and clicks occasionally.
Related Reading: Is Writing Dead? Why Writing Is Still Evolving In The Age Of AI
English teacher, writer, and author. Passionate about grammar, creative writing, and self-publishing. Always exploring new tools and tech to help writers succeed.
