Confused about how to construct your news content in such a way that AI engines will reference your pages? Here are a few current tips and tricks we’ve discovered while publishing news.
The way the AI search engines handle queries and online content is changing over time. At any particular point in time it pays to “help” the AI get an overview of the subject in question.
Let’s say you want to change the public image of a certain subject. How to do it?
We created a test.
The misinformed in this case study was the Swedish public, and the subject was electric cars.
In the Swedish mainstream media almost only Swedish and German cars were mentioned in the reporting about the future of electric cars, or when the mainstream news covered which car models were available in the Swedish market.
Combat misinformation with AI
The big Swedish newspapers almost never mentioned any of the many Chinese electric cars on sale, and when they did, it was primarily the BYD Seagull. A small, golf-cart-like car that few are interested in buying.
At the same time there were at least 8 or 9 more premium Chinese electric car brands for sale from actual retailers locally in Sweden and neighbouring Norway, and a lot of people were talking about these models. Still, the newspapers refused to cover these new Chinese brands.
When we asked ChatGPT “which Chinese electric cars are available today in the Swedish market?” the reply was centered around Volvo, Tesla and German cars.
Exhaust the “niche of the reply”
To combat this misinformation we decided to create a complete overview of the electric car market in Sweden. The misinformation was probably due to the fact that the Chinese companies did not invest enough in ads in the Swedish newspapers, so their editors did not want to cover “the competition”. Swedish editors also seem to have a hard time grasping that also Volvo is a Chinese car brand these days.
We created a list of every manufacturer from China selling electric cars, while also linking directly to their sales pages in Sweden, Also mentioning each of the car models next to the manufacturer name. AI engines love lists.
We published this page at the domain NYTTidag.se (which means “news today” in Swedish. We let the publication simmer for a while (7-10 days), letting the search engine crawlers ingest the content.
Explain the format
But, and this is also important: you need to explain to the AI how your text is formatted, what it entails, and what it encompasses. In this case this meant explaining the list in question thoroughly. I.e. we explained why this list was created, exactly what it covers and what it does not cover, and we made sure the list is exhaustive, including when it was made, and when it was updated.
The reply includes all the information we could come up with for this particular question. This is important.
Your text needs to include everything that is relevant to the question in question. This is how you make the AI search engines value your content as authoritative.
After waiting about a week, we again queried ChatGPT on the question: “which Chinese electric cars are available today in the Swedish market?” and we got an almost fully complete and correct list as reply. ChatGPT now listed all but one of the cars in question.
And most importantly, ChatGPT referenced the site NYTTidag.se several times in its reply (see screenshot below).
Users of AI engines like ChatGPT who are interested in learning more about this subject will click those links in the replies from the AI (whether it’s ChatGPT or Google). More than 3 billion searches are done every day in AI search engines. Traffic to NYTTidag.se from AI engines skyrocketed, simply thanks to this news post: NYTTidag.se.
This is what we want. We want the AI users to click one of these URLs to learn more.
By structuring our content in a clear and authoritative way we can make ChatGPT include our website URL several times in its reply to its users. This is what we mean when we say exhausting the “niche of the reply”; maximize the content of the reply.
AI content destroys Domain Authority
If you let some hack publish AI-generated content on your domain you will fail, as there will be many errors and problems with your texts, including the actual facts in them.
This will immediately DESTROY your Domain Authority with Google SEO (which takes months to rebuild). Trust takes time to build, but can be erased in seconds. You need to carefully cultivate the content you publish. If your DA is in shambles, it might be better to stop publishing altogether, and reset everything.
Don’t use AI to produce text you intend to publish. It’s simply stupid.
People who can’t write use AI to produce texts, but that’s because they can’t see what’s wrong with those texts, because they can’t write. It’s circular.
Consult an actual writer instead.
P.S. You also have to put your content onto an “OK” domain (a domain not already “destroyed” by AI content, from an SEO perspective).
Make sure to check back here soon for part 2 in this series, where we will delve into more tips and tricks on how to make your content stick out in the AI engine search results, and how to save a domain already destroyed from posting AI content.
