This is great. I agree that Health seems dangerous, but also it kinda seems like a huge opportunity for Google. Google may want to become the health expert we all turn to when we have a rash or whatever and there's a lot of bad information out there. High risk. High reward I guess.
Great article Kevin. Usefull insights and a lot of work to do for us in The Netherlands. For now we don't have Google Overviews yet. But therefore some extra time to prepare ;).
Solid! We really needed this kind of summary, thank you Kevin!
Frankly I am amazed that NONE of these 19 major studies looked into whether AI Overviews results in *more total searches*. Google claims that it does -- why aren't we looking into this? (maybe I missed it??)
Hypothetically, if AI Overviews cut CTR by 1/3rd, but result in 4x more searches, we'd net MORE website traffic. I doubt that's the case exactly, but it's a massive consideration that we need to look in to if we want to understand AI Overviews and their impact on search behavior.
I think the tricky part about looking at whether AIOs result in more searches is that we only have search volume, which is a very flawed, incomplete metric. I think qualitative data could reveal more here, and the further we get away from when AIOs launched, the harder it will be to compare the before and after.
However, we could start by looking at search volumes for keywords and related keywords that show AIOs over time.
But what about clickstream data? Compare the search behavior of sessions which have AIO appear vs. those that don't. For example, it sounds like you have a connection with Similarweb? :) Or maybe Rand Fishkin could tackle this with Datos.
G's search volume stats would not show us this. Search volume stats from non-G sources may, such as SEMrush's...
This is really useful Kevin. I've long felt AIOs are a bit of a brasher featured snippet. We increased our health content output over the last year and noticed the unusual increase in health specific AIOs (also felt bizarre to me), but I think you're probably right. They have confidence in the training data.
This is great. I agree that Health seems dangerous, but also it kinda seems like a huge opportunity for Google. Google may want to become the health expert we all turn to when we have a rash or whatever and there's a lot of bad information out there. High risk. High reward I guess.
thanks! And I agree with your point. Curious to see how that will affect health publishers.
Fascinating read!! I would love to hear your opinion on my latest article, tackling the idea of AI taking our jobs….
Great article Kevin. Usefull insights and a lot of work to do for us in The Netherlands. For now we don't have Google Overviews yet. But therefore some extra time to prepare ;).
I think it's important to mention that AI overviews are not displayed in the EU.
sure! covered that indepth in my other articles, but good to call out here as well.
Solid! We really needed this kind of summary, thank you Kevin!
Frankly I am amazed that NONE of these 19 major studies looked into whether AI Overviews results in *more total searches*. Google claims that it does -- why aren't we looking into this? (maybe I missed it??)
Hypothetically, if AI Overviews cut CTR by 1/3rd, but result in 4x more searches, we'd net MORE website traffic. I doubt that's the case exactly, but it's a massive consideration that we need to look in to if we want to understand AI Overviews and their impact on search behavior.
Thanks!
I think the tricky part about looking at whether AIOs result in more searches is that we only have search volume, which is a very flawed, incomplete metric. I think qualitative data could reveal more here, and the further we get away from when AIOs launched, the harder it will be to compare the before and after.
However, we could start by looking at search volumes for keywords and related keywords that show AIOs over time.
Absolutely, search volume will not get us there!
But what about clickstream data? Compare the search behavior of sessions which have AIO appear vs. those that don't. For example, it sounds like you have a connection with Similarweb? :) Or maybe Rand Fishkin could tackle this with Datos.
G's search volume stats would not show us this. Search volume stats from non-G sources may, such as SEMrush's...
Good point! Let me reach out to Similarweb to see what we can do :).
This is really useful Kevin. I've long felt AIOs are a bit of a brasher featured snippet. We increased our health content output over the last year and noticed the unusual increase in health specific AIOs (also felt bizarre to me), but I think you're probably right. They have confidence in the training data.
Like this meta analysis format too.
thanks a lot, harry! more to come :).
(I might steal the format at some stage, I hope you don't mind)