An array of traditional publishing organizations have signed licensing deals with OpenAI, including Vox Media, The Atlantic, Dotdash, News Corp, and the Associated Press. (We’re tracking them all here.) Meanwhile, publisher trade org WAN-IFRA announced last week an accelerator for 100 news publishers to help these newsrooms fast-track AI adoption and implementation to create content.
Some of this is defensive: Publishers know that they must work with OpenAI to ensure they’re fairly compensated for their content, and that they may find they’re able to monetize casual readership through these licensing agreements.
For marketers, however, the equation is a little less clear. Marketers, in stark contrast to publishers, are desperate to have their work surfaced by ChatGPT or Google AI Overviews. Google claims that early user feedback shows that people see AI Overviews as a jumping-off point to web content, and that clicks to webpages are of “higher quality,” with people staying on the page for longer because it’s actually attuned to their needs.
“While AI Overviews are powered by a customized language model, the model is integrated with our core web ranking systems and designed to carry out traditional “search” tasks, like identifying relevant, high-quality results from our index,” the company said. “That’s why AI Overviews don’t just provide text output, but include relevant links so people can explore further. Because accuracy is paramount in Search, AI Overviews are built to only show information that is backed up by top web results.”
That makes sense. And for marketers, that means getting picked up by AI Overviews may be a very, very good thing. Execs I spoke with now say they’re trying extra hard to ensure their SEO strategies are up to date to maximize chances of getting “picked up” by AI Overviews, trying to goose the results so their work is cited up high. Some are even pivoting content itself to ensure it’s more SEO-friendly. “Now is the time to make sure that you’re focusing on foundational [search optimization] elements, especially on the technical side,” Amos Ductan, senior VP of search at digital agency Razorfish tells AdAge. As Ductan points out, a Google AI Overview is a lot less real estate than the traditional 10 blue links – which makes search a lot more competitive.
But things are probably going to continue to get more complicated. Depending on what you ask ChatGPT, Google, or other LLMs, brand names often turn up. Jellyfish’s Jack Smyth has been researching the contexts in which LLMs mention brands. This is hugely important: As millions of users conduct queries, the answers they get from chatbots have the power to transform consumer perception.
It behooves brands to ensure they have at least some say in this. At this point, companies aren’t able to meaningfully change how LLMs describe brands and in what context – so the current hope, says Smyth, is that the content brands publish on the Internet will hopefully reach those models and keep brand perception consistent. A new metric appears to be taking shape, one Smyth calls “share of model”: How brands are perceived by LLMs, and how marketers can track, and sway that.
So when will the LLM-brand deals happen? It feels inevitable that marketers will at some point have to pay to “advertise” within chatbots, or at least safeguard against showing up in ways they don’t want to. Brand safety takes on a new import in the era of AI, and the very definition of SEO changes.