Most of us have no clue where we’re sitting with the LLMs.
If people get the answer from the AI overview, they never make it to my site.
AEO stands for Answer Engine Optimization. It is how your blog gets cited by ChatGPT, Gemini, Claude and Perplexity.
Gemini, September 8, 2026. The question was: Who is the authority in audit automation for food blogs?
I find myself using LLMs more than I’m using Google Search to answer my questions. So this is something every single person needs to take very seriously, if this is the business they wanna continue to stay in.
Pinterest right now is inundated with AI, and Pinterest has done very little to nothing to reverse this trend, period. The AI is eating up a lot of the impressions, and the creators who are spending the time, effort, and energy to produce beautiful pictures are getting snubbed by AI and are watching their traffic go down considerably. With Google, AI overviews right now are on fifty three percent of my recipes. The way we monetize the site is through ad revenue, and the ad revenue fluctuates based on the amount of traffic that’s going to the site, period. So AI overviews have hurt us significantly in the pocketbook, because they’re answering the questions that viewers would have gone to food blogs for in the past.
I ran 30 food blogs through ChatGPT, Gemini, Claude and Perplexity, eight questions each written from the blog’s own latest posts. 952 answers. 24 of 30 blogs were named at least once. 6 of 30 were never cited by any engine. The typical blog was cited on fewer than 1 in 10 of its own questions.
Read the full study, with the numbers by engine and who got cited instead.
What I did on my side is I optimized every post for AI. So we have schema, and we have added AI buttons to each post. We have optimized the format of each post so that it’s easy for AI to read it and recommend it. I created a stack with Claude after an audit we had on the site that updated every post we had through AI. It was a blessing to have all of the posts updated by AI. Unfortunately, I got really aggressive in updating posts and did over nine hundred posts in one month. Google thought that was way too much, which it was, and put me in Google jail, period. I know I’ll come out, but this was a lesson learned that I’ll make sure never happens on anybody’s blog, period. I make the mistakes on mine and learn from them and make sure I don’t do them on another blog.
The AEO add-on does five things.
Number one, it scores your site. We write twenty real questions based on your posts. We ask all four, ChatGPT, Perplexity, Gemini, and Claude, and record who they cite. We give you a score out of a hundred, and the names of the sites winning the questions.
The second thing is the doors. We check that every AI crawler can read your site. Most blogs do pass, but there are ones that fail and are invisible, and they don’t even know it.
Third is the identity. We make sure that everything is aligned with who the owner of the site is. Most blogs blur the owner and the brand together in their schema, which makes it difficult for the LLMs to have a clear picture of who they’re recommending.
Fourth, the monthly recheck. Later in the month, we run the same twenty questions and measure our progress on what the search engines say when asked about your brand.
And fifth, and most importantly, and which a lot of sites don’t do, is we provide the fixes. Most products I’ve seen will identify how you’re showing up in the LLMs, but they don’t show you how to fix your posts so they show up in the LLMs. That’s where we come in. We offer that solution so you will be cited in LLMs.
Before we do anything, we make sure that you are crystal clear, through our checklist, of everything we’re doing on your site. If you don’t check it, we don’t do it. We wanna make sure we are as transparent as possible and everything being performed on your site is as crystal clear as possible. We would never do anything without your permission.
This is for every blogger, because most of us have no clue where we’re sitting with the LLMs. And if we do know, how do we fix our blogs to make sure we’re getting cited? None of that information is out there. And it’s so important in this age of Google AI overviews to diversify where your traffic’s coming from. I see LLMs taking more and more traffic away from Google search.
The honest version.
We are still building case studies for this add-on. We will not quote you a traffic number we cannot back up, and you should be careful with anyone who does.
What we can promise is a measurement. You get your score today: twenty real questions written from your own posts, asked across ChatGPT, Gemini, Claude and Perplexity, with the names of the sites winning them. Thirty days after the work ships, we run the same twenty questions again and show you exactly what moved. Nobody else selling this offers a measured before and after.
Pricing for the add-on is quoted after your free score, because the work depends on what the score finds. If you have never had an SEO audit, we will walk you through what a real one covers and can introduce you to the auditor we work with. If you already have one sitting in a folder, we will read it and come back with what it would cost to actually do the work, in hours and in dollars, by hand versus through our system. We know that number from our own site: a 33 page audit on 1,500 recipes, quoted at $70,000 and about two years of a VA doing two posts a day. We ran it in under two months.
If someone wants this product exclusively and doesn’t want to go through the audit because they never had one performed for them, then the first step would be the free score. They would see where they’re ranking in LLM search, seeing where their vulnerabilities are, so they can make the best decision for their business moving forward.
Four fields and one question. Results every Monday.
Is ChatGPT citing your food blog? Find out.
Four fields and one question. We run scores in a batch every Monday and email results the same day.
This is an add-on to the Audit Automator. The other add-on is Get Published. Our own results are on the proof page.