The New York Times recently reported some of the commotion around Google Magi and the tech giant’s attempt to catch up with some of the developments by Microsoft who introduced AI to their Bing search engine.
Naturally anything to do with big changes to search gets the SEO/PPC community whipped up into a bit of a frenzy. This may well be the biggest paradigm shift that has happened in recent years, but it’s not the first time that search engines or Google have changed how things work, and it strikes me that much of the best practice around search engine marketing will remain the same.
Google Magi-powered search will still rely on websites, so the same principles will still apply
Here we’ll look first at the SEO situation with regard to Google Magi, and others.
Let’s remember that an “AI” is, in most cases when we’re talking about information retrieval (which is what a search engine is one example of), just a cleverer way of interacting with underlying data.
Through additional information such as personalisation based on search history, browsing habits, geographical location, device etc. one could easily argue that Google and other search engines are in fact a hybrid using text to match queries to results and recommender systems, which broadly use contextual signals to optimise the response.
Behind the scenes, Google is still going to be collecting data from websites using its various crawlers, and SEOs will still be expected to make sure that content is well themed, fast and with the appropriate markup (including the use of microformats to help identify things like events, recipes etc.).
Google and co have already been on a mission to get people using these formats so they can index them directly as recipes, books, films, games etc. and serve elements of that content directly on the results pages. Their plans for Magi seem to suggest that the chatbot element may help narrow someone down to an even more precise result (rather than the standard positions 1-10, 11-20) more quickly through the chat experience. We all know that you’re not really getting much traffic if you’re not in the top 3 or so anyway, dependent on vertical and the volume behind it.
What traffic does get sent direct to a website is likely to prioritise the same things that it does now — user experience and backlink profile (I’ve simplified things there a bit), and certainly the first of those should be a priority for any website regardless.
A final note on this, people will not automatically start to chat to a bot. They will search the same way they used to, and “traditional” keywords will likely still be important, because they are still going to be the most efficient way to satisfy a user that knows what they want, and they are the ones who are most likely to convert. It would be a huge own goal if the experience was so indirect that it became less efficient for a large swathe of search engine users, and they turned to more “primitive” search engines as a result.
Google will still need ad revenue
They can’t change their model fundamentally, or too quickly, for risk of losing out on lots and lots of ad revenue whilst agencies or brands try and figure out how to keep up with them. They will have to enter into a long process of testing and consultation, not just with their external stakeholders but also the teams that run Google Ads internally.
And let’s not forget that Google Magi isn’t released yet. They may be scrambling tons of devs to work on it, but if it’s still being built then we are probably months from any substantive changes.
It does sound from the NYT article that there will be changes to how transactional the Google experience is. That could play into the hands of bigger brands and agencies using Google Shopping already — and could be a nightmare for those smaller companies with no Google Ads budget or limited tech resource.
Prediction #1: Google Magi isn’t really going to change that much for DR marketers
In conclusion, I think that the “head terms” that people talk about in search engine marketing circles are going to be safe. Google can’t afford to lose advertisers. Anything overtly brand-related or transactional doesn’t need a conversation with a chatbot to come in and make the user journey laborious.
Imagine the situation: you search for “men’s trainers” and get shown a bunch of ads and 1 or 2 organic results for men’s trainers. You say “actually some red ones” and you get shown the equivalent of the top of a SERP for “men’s red trainers”. You add “for wide feet” and you get exactly what you’d expect. What’s the difference, apart from how the information is presented? The idea of a search session becoming more and more focused as users head down the funnel refining their own queries is not exactly new.
Prediction #2: the winner in the chatbot-powered search engines will be the one that runs the fastest
I just tried the above out on the Bing AI, and it was so slow to respond for these basic searches that it made me laugh. We’ve been told that search engine marketing is all about the Zero Moment of Truth and not keeping consumers waiting — if that’s the case then any search engine that leads with an interface this appallingly slow is going to shoot itself in the foot.
Conclusion: calm down a bit everybody, for a moment
I know that this post is going to end up being one of those with comments like “this post aged well” but I’m putting it out there anyway. The takeaway points are:
- Websites won’t disappear, the Google index and most large language models rely on websites
- PPC ads won’t disappear or change fundamentally, because Google depend on adverts
- Best practice for SEO will stay basically the same, especially the tech side of things
- Transactional SEO may become harder if more results and transactions take place on Google Magi directly — that would likely play into a PPC marketer’s hands through Google Shopping
