This week’s #SEOForLunch sponsors are Semrush and Jolly Search.
Last week at Tech SEO Connect, a speaker argued that judging AI search by clicks and revenue was a mistake. I made a similar case a few weeks ago. I still believe AI search has value as a brand channel even when nobody clicks. I also know exactly what happens when I walk into a leadership meeting. Somebody asks how much money it made.
So what do I actually show them? I call it SEO+ Reporting.
Hurry up and sit down. Leave your vegetables in your locker.
AI transparency note: AI helped outline this issue, draft and edit sections, and generate the hero image. The perspective and client experience are mine; I approved the final version.
Thank you to Semrush for sponsoring this week’s #SEOForLunch
Meet the new AI visibility data standard
As prompt databases get artificially inflated, it’s time for a reset in what we consider trustworthy AI visibility data.
Most important is ensuring data it’s part of a connected context. Semrush’s AI visibility data is led by real customer demand, follows a topical structure, and is AI response-validated.
Start With the Money
I’ve written before that leaders want AI reassurance, not a lecture. They might be genuinely curious about how AI search works. But when they’re deciding where to put next quarter’s budget, “it depends” only gets you so far.
If your leadership team talks, dreams, and farts dollars, start with the dollars. In the dashboard below, I show total revenue for the initiative first, then break out revenue and sessions by channel. AI assistant referrals get their own line right alongside organic search, paid media, email, and the rest.
I built this report for a specific client initiative with help from AI. I’ve blurred client details and figures so you can see how the report is laid out.
The AI line will probably look small. Cue the inevitable question: Is the website broken, or does AI just hate us?
Before I answer that, I show where the numbers came from. In GA4, I report sessions and purchase revenue for the selected period. I show GA4’s AI Assistant channel alongside the custom grouping I’d already built for known referrers like ChatGPT, Claude, and Perplexity. Those views can overlap, so I keep them separate.
Introducing SEOplus Reporting
You’ll be shocked at how similar this looks to your existing SEO reporting. This is because it is similar, but I’ve leaned into the limited official data we have available on AI search.
Most notable: GA4 attributable sessions/revenue, GSC URLs/Impressions, and BWT citations/mentions.
I share an example of a specific report I built (yes, with the help of AI) for one of my clients for a very specific initiative they are focusing on. Out of respect for my client (and my signed NDA) a lot of the data is blurred out but the layout and metrics themselves are what matter for this post.
Here’s Where the Plus Comes In
At this point, you might be thinking, Great, Nick. You made an SEO report. Yep. Now we can answer the question GA4 leaves open. Where are we showing up in AI search, and which pages are getting picked up?
Google Search Console shows impressions in Google’s generative AI features, including AI Overviews and AI Mode. I track how many distinct URLs received impressions and break those impressions down by page, country, and device. Google’s report documentation
Bing Webmaster Tools shows citations, cited pages, and sampled grounding queries. Those queries give me a sense of the topics Bing associates with the content it cites. Bing’s report documentation
I put the trends next to each other so we can see how each moves over time. Google impressions and Bing citations measure different things, so each gets its own scale. I also label the dates because the GA4 cards show totals for the selected period, while the visibility cards show a single day.
This is where the report gets useful beyond the leadership meeting. Say Bing is citing your warranty pages while the buying guides you’ve invested in barely show up. Now you know which content is being picked up and where to dig next. And if Google impressions rise while tracked assistant referrals stay flat, you can show both trends without assigning a dollar value to the gap.
“Not Possible” Is the SEO Loser Mindset
I know that phrase pisses some people off. I’m okay with that. There’s a real attribution gap here. There’s also enough first-party data to put something better in front of leadership than a shrug or a made-up AI revenue number.
Give them the revenue you can track. Show them where the brand is turning up. Tell them what you’d investigate next.
You can spend the meeting explaining why the report they want is impossible. I’ll be shipping mine and collecting checks.
~Nick
Thank you to Jolly for sponsoring this week’s #SEOForLunch
Can AI Even Read Your Website?
Brought to you by Jolly Search.
Quick test: open your homepage, view the page source, and search for a sentence you can see on the page. If it isn't in the source, ChatGPT, Claude, and Perplexity can't see it either.
We've audited sites that look perfect in Chrome and completely blank to AI crawlers. The owners had no idea. The Sitewide AI Visibility Audit tells you:
Whether AI can read you: 9 technical factors, graded A to F.
What to fix first: findings ranked by impact, with a plan your developer can run with.
Where AI is looking instead: your presence across the channels AI engines actually cite.
$250. Five business days. Credited if you start any Jolly plan within 30 days.