@tjrobertson52

As people are starting to use large language models like ChatGPT to find businesses, US marketers are realizing that it's much harder to track the results than it was with traditional SEO. Really, US SEO's have been kind of spoiled with how much data we get. We know exactly what people type into Google and which webpage they click on.

In large language models, we have no idea what they typed. And even when they do choose to work with us, they typically aren't clicking through to our website so we can't track it at all. It's not too dissimilar from billboard or radio or magazine ad campaigns.

there are tools that help us gauge visibility across likely prompts and we can extrapolate from the limited traffic that does come directly from the large language models. we can get a fairly accurate view of if things are trending upwards or downwards. And the tools can tell us the most common sources and types of content that large language models are citing.

So actually, I take it back. It is much better than billboards or radio. It's just not what we've been used to as SEO's.

We don't know what people are typing into these large language models and we don't have an accurate number of how many people are finding us through them. However, this could all actually work to your advantage since your competitors also don't have that data. Most of them are going to devalue the importance of showing up in large language models, they're going to lean on marketing channels with reliable conversion data.

And as a result, they're going to underspend on what I believe is becoming the most common channel people use when deciding which brand to work with. So my prediction is companies that are willing to make the investment on limited data are going to see an outsized return on that investment.