When we first heard about the Google integration of Generative AI, we quickly felt that it would radically change the Internet. After writing a detailed explanation as well as already thinking about the implications for SEA, we conclude the triptych with the talk that started it. SEO-Marketer Jelle gives an update on the latest developments. This brings him and Cecile (digital illiterate/doomsayer with writing talent) back to the question that made us come up with this triptych: is SEO still relevant if Google automatically generates the ideal search result?
SEA marketer Jannes said it before: we are reinventing the Google wheel. According to Jelle, we need not fear that these changes will cause SEO to disappear from the scene: "Big names from the SEO world have already spoken out about this, they are not worried about replacement. In any case, the work is going to be turned upside down. We don't have to fear replacement, but we do prepare for change."
Cocktail of existing information
According to Jelle, the angle of prediction must shift. An AI always needs human input in one way or another. "Generative AI is a cocktail of existing information. Is it right to focus exclusively on what happens on the front end in the search engine? Or should the conversation focus more on the SEO marketer's efforts to contribute to good search results?" In that respect, Generative AI is an interesting development for SEO marketers: good copy becomes more important. You're no longer competing for that top search result, but for a namedrop in the AI-generated result. As long as that connection between AI and existing content is there, SEO is relevant.
Optimism, pessimism and a lot of smokescreens
Here, of course, we are assuming that the integration of Generative AI will proceed exactly as planned by Google a few months ago. The testing process, which has recently been extended to India and Japan, shows that a lot is still possible. Cecile gives an example: "We saw in the beginning of testing that Google's ideal search result first was a running text."
Source: Google.
That text gradually changed into a list of short descriptions, sorted with bullet points. "The most interesting change, though, was that links popped up next to those bullet points: suddenly it started looking very much like a list of ten search results again." In other words: back to the old Google.
Source: Shalom Goodman, Twitter/X.
Two weeks later, the bullet points had been replaced again with running text. The links disappeared again and were replaced by drop-down menus that you had to click on to see resource links.
Source: Search Engine Land. Also check out this GIF (Google).
Those drop-down menus turned out to be optional: the user must turn on the visibility of that source attribution himself. As a copywriter, Cecile worries about this: "Who turns on the source attribution? Do people still click through to the original content that Google links to or do they get stuck with the summary that Google creates? Consider Wikipedia, for example: many people start on Wikipedia and don't search further. They usually don't click on the footnotes. If something similar happens at Google, they don't even reach your landing page."
Jelle emphasizes that this example mainly shows that a lot is still uncertain. "We don't yet know how much page space such an ad takes up. If it's a small ad, we don't need to panic." Perhaps, as an SEO marketer, you continue to participate in the competition to be the first search result, and it's now all about being the top source ad in Google's AI-generated response.
That format of the summary may determine how Google uses your content, because you have no control over that. If you search for Tomahawk in the future, Google will generate a summary for you based on our website. Whether the unpredictable combination of different tufts of text from our website is logical or at all correct is up to Google. Cecile: "I don't deny that I secretly fear a Frankenstein's content monster made up of all sorts of snippets of text stitched together that I wrote for completely different purposes."
Here's how to claim the top spot
How do you get that top position - no matter what that top looks like? The solution is in long-tail keywords: keywords with low search volume. It sounds contradictory to bet on those if no one is searching for them. In the future, that's exactly what makes the difference: why would someone google "pumps red" when you get a better result with "red pumps with 10-centimeter block heel and 2-centimeter platform"? By using long-tail keywords, you emphasize what makes your product unique, so you reach those who search for it faster.
The pitfall here is keyword stuffing: incorporating as many long-tail keywords as possible into your content so that your content always fits the search query somewhere. Compare it to writing long texts when you had a test in school: if you give all the answers, you will give the right answer somewhere (of course, no one at Tomahawk is guilty of this...). Seasoned Internet users recognize this tactic, and search engines are not happy about it.
According to Jelle, it's better to do the opposite: "You look for the niche in which you make your content special, so that it gives the one right answer to that one detailed search question." Unfortunately, with this strategy, you run the risk of not getting a feature in the search results if the search query doesn't quite fit, because someone else has that better fit to catch.
Jelle emphasizes that an SEO marketer must make the trade-off between hyper-specific or broad content, including all the benefits and risks involved. "Keep in mind that Google's goal remains to provide the user with the ideal search result and that in-depth queries will prevail in the future. There is no one-size-fits-all solution that applies to all search queries."
Attribution is still a thing
Because we don't yet know how Google will design the ideal search result, we don't yet know how revenue will be attributed. Jelle: "Is the ideal search result organic, or the result of a paid ad?" Distinguishing what someone is paying for or not is essential in The New Googling. Time will tell how attribution is determined. Jelle predicts that the understanding of how a Web site ranks will become clouded. The box with the ideal search result seems to have no set formula for success to anticipate as an SEO marketer. Is that a bad thing?
Frankenstein's content monster is real?
That's not so bad. The competition is just as much in the dark as you will be, which levels the playing field. Jelle: "That presents cool challenges and lots of opportunities. This keeps us as SEO marketers innovating and discovering new ways to make our content relevant all the time."
Cecile wonders if she is overprotective. "Maybe I'm overestimating how smart the AI is going to be," she says. Currently, Bing's AI is pretty similar to Google's big plans. Jelle shows his colleague a search result he created with Bing's AI (pictured below). Jelle: "If these are the results that the Google AI is going to come up with, you, as a user, have no use for them." Nowhere on our website do we mention that Tomahawk provides conversion optimization, UX design or website development. We don't offer this, yet it pops up in search results.
Bing quotes our website as the source: how can that be? The incorrectly interpreted information Bing pulled from this page. The content from Tomahawk that appears in the search result is randomly chosen, interpreted differently and puts the reader on the wrong track. This is similar to Frankenstein's content monster mentioned earlier. How do you fix this? If the risk of interpretation errors forces a copywriter to avoid certain words, that's an undesirable situation. That's why we're going to have a heart-to-heart with Bing's bot, so that the AI interprets our website better next time. That way, we keep the final say on our content.
