Insights
One Year Of AI Overviews
AI Overviews now appear on approximately 48% of tracked search queries, up from 30% a year ago.
That's one of the data points from BrightEdge’s look at how AIOs have evolved over the past 12 months.
But, it appears BrightEdge is pushing back against the panic some marketers feel about AI "destroying" SEO. They're saying: "Yes, AI is important and growing, but don't let the 48% distract you from the 52%."
When a feature appears on nearly half of searches, it feels like it's everywhere. But mathematically, the absence of that feature on the other half is equally significant.
Focus on the 52%
So, let's focus on the 52% of queries that don't show an AI Overview. This means traditional organic rankings remain the primary experience (just) for most searches.
It gets even more interesting when you dig into the citation data.
Only 17% of AIO sources also ranked in the organic top 10. So even when AIOs appear, traditional rankings still matter for the actual clicks. And when AIOs don't appear, organic results remain the entire experience.
While headlines like, "AIOs Grew 58% in 12 Months" attract attention and may set off alarm bells, we need to remember that we're not facing a binary choice between "old SEO" and "new AI optimization."
We're in a messier, more challenging place where both matter, where the relationship between them is still unclear, and where the fundamentals are probably still your best bet.
Your industry changes everything, too. Get the numbers and observations on AIOs in this BrightEdge report.
Does answering PAA questions correlate with top rankings?
Here's something to think about.
AlsoAsked published a study showing the relationship between answering "People Also Ask" (PAA) questions and ranking in the top Search positions.
Let's address the math first.
In SEO ranking studies, we often see negative correlation coefficients.
A lower ranking number is better than a higher one. A negative correlation (where the correlation coefficient is closer to -1.0) indicates a stronger relationship.
It means that as the number of questions answered on a page goes up, the ranking number goes down (closer to the top spot).
The study found that this correlation is most significant within the Top 5 positions in Search.
Beyond the first page, the correlation drops sharply and can even reverse on Page 2, suggesting that PAA optimization is primarily a signal for top-tier rankings.
The numbers
1. Strongest Correlation: Transactional Intent (-0.9573) For pages where the user is ready to act, the link is nearly perfect.
This indicates we should prioritize PAA coverage on bottom-of-funnel conversion pages first, where completeness matters most to Google.
2. Very High Correlation: Commercial Intent (-0.8941) When users are in the "consideration" phase, the pages that answer the most auxiliary questions tend to win the top spots.
3. Strong Correlation: Informational Intent (-0.8867) Surprisingly, this was lower than commercial intent.
The hypothesis here is that for informational queries, specificity may outweigh breadth. Answering one or two highly relevant questions deeply may be more valuable than answering ten general questions.
4. Weaker Correlation: Navigational Intent (-0.7414) While still relevant, the relationship is weakest here because Google prioritizes the official brand or destination regardless of the volume of questions answered.
Understanding these correlations is the first step. The second is realizing that simply having the content isn't enough because the quality of the answer becomes the deciding factor.
This study gives you a framework for your own experiments and also notes how the methodology could be improved.
Will you experiment with this?
My questions about ChatGPT ads
While we're all hoping to catch a ChatGPT ad in the wild, a LinkedIn post got me thinking about how users will respond to ads in LLMs.
A carousel from SE Ranking features the thoughts and reactions of SEOs and PPCers on OpenAI’s move into paid prompt results.
One area that seems underexplored is the long-term behavioral impact of ads in LLMs.
How might subtle exposure to sponsored content shape user decisions over time, even if the ads are clearly labeled?
Advertising that is poorly integrated or irrelevant could erode credibility and I suspect users will be skeptical of OpenAI’s ability to control that.
Then there is the competitive dimension.
If one platform maintains a cleaner, ad-free experience, could that become a strong differentiator?
Sometimes, choosing what not to monetize becomes the strategy.
OpenAI is clearly testing ad integration by only showing ads in the US for now. We'll wait for the early feedback from users (and anyone trying to get an ad to trigger!).
Do you believe ads will be a success in ChatGPT?
Hit reply with a 'Yes' or 'No" (and send me screenshots if you've seen ads appear).
Just in:
Perplexity is stepping away from advertising over concerns that it will erode user trust.
"A user needs to believe this is the best possible answer, to keep using the product and be willing to pay for it," a Perplexity executive told the Financial Times.
Do you follow me on LinkedIn? I share regular tips and stories I don't have room for here. Come and join me.
