On May 15, 2024, Google integrated AI into their search results for everybody.
What this means is, if you head to Google and search for something like “anxiety,” you may now see an AI overview at the top of the results, which for me just defines the term.
We’ve known this integration was coming for quite a while, ever since Google announced the Search Generative Experience in May 2023. We can think of it as a kind of rival to OpenAI’s ChatGPT, among others.
Now that Google’s AI is here, we can say it’s… kind of a letdown, which we detail below.
Of course, we are mainly concerned with how Google’s AI affects Private Practices. The answer to that is simple: for now, the effect is null. The future is a different story.
How Does Google's Search AI Affect Private Practice Websites?
Right now, the keywords that potential clients use to find Private Practices have not been hurt by Google’s AI integration. (We say this tentatively.)
Keywords like…
- Anxiety therapy
- Depression counseling near me
- Trauma therapy
…don’t trigger Google’s AI overview.
What does trigger the overview are informational searches, like “depression.” These are the kinds of searches that, in the past, would lead you to a national website like the APA or Wikipedia. Now, they sometimes show AI-generated results.
Interesting enough, “trauma therapy” is a local keyword, whereas “therapy for trauma” now gives us an AI result without any GBP results. This means Google treats the former as a local search term and the latter as a national search term despite their semantic similarity.
What Google treats as national (and therefore, AI-worthy) and local seems to lack an established logic.
The main way this will affect Private Practices is through blogs. Those long-tail keywords for blog posts, such as “How anxiety can create false memories,” will soon be answered by AI summaries (so we predict).
Of course, if you follow our strategy, you know blogs are not there to rank highly on Google, though it is always a benefit when they do. Blogs exist to scaffold Specialty Pages as part of the silo structure.
What Do The AI Results Look Like? How Do They Work?
Let’s return to our previous example. When we Google “depression,” we get what’s called an AI overview at the top of the results.
(Note: search results are unique to each person, based on your IP and what personal data you share with Google. You may not get the same result. Through testing, we weren’t always able to get Google to generate AI results for the same keyword. It appears there are still quite a few kinks in the machine.)
If we click “Show More,” we see the full generation.
The way the AI works is, Google pulls information from various—arbitrarily—trusted sources and then collates that information into a single summary. If we click the carrots (or down arrows), we can see which websites they’ve pulled this information from:
- World Health Organization
- Cleveland Clinic
- SAMHSA
- MedicalNewsToday
- Mayo Clinic
All sites that are quite large and safe for Google. What you’re unlikely to find are Private Practices in the results—at least for a query like “depression.”
Admittedly, Private Practice websites don’t rank for those general terms anyway.
How Accurate Are Google’s AI Results?
About as accurate as any other AI—so, results vary.
From now on, when searching for complex questions (anything with nuance), be wary of AI-generated results.
In the first 24 hours of Google’s AI release, the AI said, among other things that…
- Drinking urine cures kidney stones
- Mischaracterized the four horsemen of the apocalypse
- Claimed that food names that end with “um” include “Applum, Banana, Strawberrum, Tomatum, and Coconut.”
Google seems to be on top of every viral mistake their AI makes, and they quickly rectify it (our guess is they manually fix the AI results). Shareholders are watching the rollout, after all. Though that doesn’t mitigate the AI’s confident stupidity.
So far, hallucinations seem to be baked into the nature of these learning models.
Which is one reason why Google’s help forum has been crammed with people complaining about Google’s AI. One Product Expert even recommended using another search engine. And the reply itself is suspiciously algorithmic…
The other reason being that the AI overview, which is inaccurate at times, pushes website results further down the page. Meaning, that search for many queries has become worse.
One way to look at it is… being wrong is the one fact about current AIs that make them seem human. But AIs that reflect human fallibility don’t seem very useful either.
What Does This Hold For The Future Of Private Practices and Google?
The immense blowback to Google’s AI integration may cause a course correction in the not-too-distant future. Though we’re holding our breath on that one until Google’s stock price slides.
For now, the keywords that connect potential clients with Private Practices are not affected. And we’re guessing they won’t be, because the search intent is different from an informational search which the AI has been employed for. Instead, local terms will continue to show directories, GBP profiles, and Private Practice websites.
But if you do find yourself searching on Google, and run into an AI result, make sure to double check the information it gives you.
