Brand mentions / AI visibility

ChatGPT now accounts for 10% of search online, at least according to the head of ChatGPT. Let's talk about where that number comes from and what business owners should be doing to get found online right now.

That number was part of a LinkedIn post discussing how well ChatGPT is doing right now. Several people in the comments asked where that number came from and they did not get a response.

I think it's pretty clear where the number comes from though. It comes from opening eyes initiative to promote themselves as the leader.

And at least right now, they really are the leading AI assistant. Even though it looks like Google currently has the smartest model, as many people have pointed it out in my comment section, a lot of people don't need the smartest model.

They just want the best product. And ChatGPT has always been focused on product while Google's always been more focused on research.

So while I do think Google is currently winning and will continue winning for the foreseeable future, ChatGPT is not going away anytime soon...

rds large language models affect your strategy? We're moving from a winner takes all economy to a situation where there can be many winners.

In traditional search, the average business is fighting over a few very competitive, very common search terms. If someone is going to Google search to look for a dentist, as a dentist, you need to be in the top three positions.

Otherwise you're gonna get little to none of that search traffic. Large language models change things in two fundamental ways.

If that same person goes to ChatGPT looking for a dentist, they're unlikely to type in something simple like dentist near me. More often, they're gonna have a full conversation with ChatGPT.

And it's the same thing with Google's AI mode. ChatGPT or Gemini will then run not one but a series of searches in their own search engine.

However, these searches are gonna be way more specific than dentists near me. It's gonna take all that context it has from your conversation as well as previous conversations and run a series of specific searches...

best strategy is to create as much content as possible, get as many people talking about your brand online as possible. As I've said before, the noisiest brands are going to be the winners.

Right now the entire SEO community is talking about what it takes to rank in AI like ChatGPT or Google's AI mode. And one common sentiment that I see keep coming up is that ranking in AI is exactly the same as ranking in Google Search.

Because of course, these large language models use search engines to determine recommendations. So for example, ChatGPT will run a search using Bing when you ask it to provide recommendations on businesses.

Of course, Gemini will use Google. Therefore, the rationale is that you should just do the same things you're doing to rank in those algorithms, because these large language models the exact same ranking.

Now that's partially true, and everything you're doing that helps with SEO will also help you show up in these large language models. However, it also misses two very important distinctions.

Is that not all output from large language models comes from search. These AI's are pre trained on all the content on the internet and therefore have a very deep understanding of big brands and already an understanding of very small brands...

players on mobile device while standing in line. No human would ever search that way because they understand that it's very unlikely there's an article with that title.

These large language models search that way all the time. It's also easier than ever to create high quality content at scale.

So that's one way SEO is changing, is it's suddenly worth it to make content addressing these hyper specific queries. And regarding the first point on brand recognition, it's now more important than ever that your brand is mentioned on the publications that AI.

Is crawling. And unlike traditional SEO, it's much less important that it actually has a link to your website.

A mention alone is enough. And these LLs are way less picky than traditional search engines.

I constantly see them citing obscure blogs and even spamy press releases distributed by the company themselves. So while we still don't know what the best practices would be, that's where my bed is right now.

And that's what I've been experimenting with...