It's no secret, I've been working on joining PPC and SEO data for 7-8 years now trying to find unique value that siloed teams miss and bring that thinking and execution to our industry and our clients.
So recently I started brainstorming with one of my in office co-workers and she came back with the most killer dashboard in Signals (our data platform that combines PPC and SEO data at a granular level).
Now we have a whole new set of data to harvest and join to PPC and SEO.
AI Overviews. AI Overviews have links, answer text, various layouts, this is turning strategy on its head. We now have these big AI powered answers sitting in boxes all around our paid ads and organic listings (the layouts will likely change a TON over the next 2-3 years).
This new layout causes a LOT of headache for decision makers...who have a new big question to answer:
How has AI overviews impacted user behavior & how has that shift impacted my thinking on the the investments I make in search?
First we know AI overviews impact click through rates on paid and organic. Here is our study that outlines those findings.
As those CTRs decrease, we know the total addressable market didn't instantly decrease. But our leading indicator data (rankings, traffic, etc) are tanking, leaving a lot of us scratching our heads on what to do.
One of my biggest challenges to level up my consulting game is how can I create a visual that I could put up on a screen, step away from the mouse and have a 30 minute strategy session. In response to this challenge, I built this analysis.
This visual allows you take an analysis, talk strategy and priority all in one visual during executive level conversations.
Analysis + Strategy + Priority - How I see it
1️⃣ - Keyword themes with HIGH % of AI Overviews & LOW Conversion Rates
Medium Impact: these words convert lower, yet they possess high AI Overview penetration, indicating that AI overviews are cannibalizing their visibility. However, these keyword themes are not big drivers of leads and sales anyway. These themes might get lower paid impressions over time, and thus you get a hit in CVR on those keywords - this is a medium level threat.
2️⃣ - Keyword themes with HIGH % of AI Overviews & HIGH Conversion Rates
Big Impact: any keyword themes in this quadrant will make challenging for you to drive conversions, full stop. AIOs are highly visible on high-converting keywords, I would start looking at other ways to gain value in this area. Ensure your brand is mentioned in these places and that your on-site content is tight, consistently updated, and accurate and I would partner with PR & Social to promote these assets (for this ecommerce client, literally many of these were about sizing, hint hint many are derivatives of branded keywords). Getting this content out in other place is the key, analyzing the sources is crucial, as these themes do drive conversions, yet AIOs will prevent those people from reaching your site. Find new places to post content to engage with these audiences, and if these pages drive don't let others take up this space.
3️⃣ - Keyword themes with LOW % of AI Overviews and LOW Conversion Rates
Low Impact: While there are not a few AIOs present, which is advantageous, these words also do not drive meaningful conversions. You can and should keep an eye here, especially on themes that move into new conversion rate territory, you don't compete with AI Overviews here so you've got a chance to attract traffic. This could assist with retargeting efforts, but keep in mind that these keywords won't directly convert. So think of retargeting, email capture, or anything else that allows you find new ways to push them down the funnel.
4️⃣ - Keyword themes with Low % of AI Overviews and HIGH Conversion Rates
Go All In: With low AIO overview penetration, your ads and organic results will get clicks AND you have a high likelihood of converting these visitors. Optimize here now! It seems like everything is moving to an AI overview world and you got numbers to meet this quarter.
How do I build this?
If you are wondering "what all am I looking at here?" Let me give you a bit of background on the data. Basically you take all your paid search terms and splice them into ngrams (I prefer unigrams but this time bigrams also create a F-ton of value). Now you have ngrams as themes combined with performance data (clicks, costs, conversions, impressions) whatever you deem important.
Pro-tip: Pick 1 metric. Don't overwhelm an exec with all those metrics pick one.
Pro-tip: Bigrams will make your dataset bigger and slower.
Now you have your paid data Ngrammed, ready for Step 2?
You take those search terms from paid, and run a Google scrape on all of them, that is how you get the presence of an AI overview - so now you have SEO data as well if you wanted to take it up a level, rankings are there too.
Pro-tip: Look for your presence in those AI Overview text to know if your brand is showing up. That would/could impact prioritization. But it could also complicate the story and analysis a ton. We seek out brands in responses in our Gen AI tracker, I didn't want to go through the headache of joining them.
Pro-tip: Look for your rankings and your urls. Do you see a drop in CTR in paid or organic on the pages that have AI Overviews showing? Just keep in mind complicating visuals can run you into circles.
You get surprises like these, which helps you make your own calls on what triggers change your desire to target keywords and themes with more or less aggression.
Now you can use Power BI, Tableau, Looker Studio to join on the search term and boom! Heck, maybe even just use an AI agent to do it all for you, I know I'm going to try to do it and replicate some of the impact we had with our first AI agent on NinjaCat.
The real goal here, in my honest opinion, is to assess risk and get your organization a POV on what to prioritize with the resources you have, not to build the world's most bulletproof analysis.
