Real human. Fast replies.
💬Text me(386) 569-1227
📞Call me(386) 569-1227
✉️Email mechenzo@honesty.im
📝Send a messageI reply same day
Bing did something genuinely useful this week, which was not on my 2026 bingo card. On October 13 it added a beta called Topics Portfolio to the AI Performance report in Bing Webmaster Tools, and it shows which subjects AI answers actually cite your website for.
Not what you think your site is about. What the robots think it’s about. Those are often two very different businesses.
The short version
It’s a new tab in the AI Performance report inside Bing Webmaster Tools that sorts the questions your site shows up for in AI-generated answers into broad topics. Instead of staring at one query at a time, you see the shape of your whole presence.
Quick translation of the jargon. When an AI writes an answer, it first runs its own searches behind the scenes to find sources. Bing calls those grounding queries. Topics Portfolio bundles yours into clusters and gives each one a few numbers: Topic Size (roughly how many citations go out across all sites for that subject), Sites Cited (how crowded it is), and Queries (how many different questions live in it).
Under the table sits Top Opportunity Queries, which is the part I actually care about. You can look back one, two or three months. The full rundown is on the Bing Webmaster Blog, and yes, the numbers are approximate. It’s a beta. Treat it like a weather forecast, not a bank statement.
Because it’s free homework for AI search, and almost nobody’s doing it. Most owners I talk to have never logged into Bing Webmaster Tools (fair, it’s nobody’s idea of a fun Saturday), so the ones who do get a head start.
Think about how people find a roofer or a wedding venue now. A growing chunk of them never see ten blue links. They ask an assistant, it reads a handful of sites, and it names a couple of businesses. If your site isn’t one of the pages it read, you weren’t in the conversation. You were in the parking lot.
Google’s been adding its own AI reporting to Search Console too, so this isn’t a Bing quirk. It’s the direction everything’s heading: tools that tell you how AI sees you. I wrote about the same blind spot from the traffic side in my post on tracking Gemini visitors. This one’s the other half of it: not who showed up, but why the robot sent them.
Usually it just means there isn’t enough data yet, not that AI hates you. Bing says the tab only appears when a site has sufficient data, and smaller sites may see a short list or nothing.
That said, “nothing” is still a nudge. If a site with five pages, a stock photo of a handshake and a contact form doesn’t register anywhere, the fix isn’t a magic setting. It’s giving the AI something worth quoting.
You make it dead obvious what you do, who you do it for and where. AI recommends specialists. It does not recommend “a little bit of everything,” and honestly, neither do people.
I see this all the time: a pressure washing site with a blog post about crypto, one about the owner’s fishing trip and a page that just says “Services” with six bullet points. That’s not a topic. That’s a junk drawer. If Topics Portfolio ever does show up for a site like that, expect the clusters to look like a ransom note.
Here’s what actually moves the needle, in plain English. Write real pages for the jobs you want, one service per page, and answer the questions customers ask in the first sentence. Add structured data, which is basically a name tag for robots: who you are, where you serve, what you sell. Put up an llms.txt file, a short cheat sheet that tells AI tools what your business is and which pages matter. Keep your name, address and phone exactly the same everywhere, and make sure your Google Business Profile matches your website instead of contradicting it.
None of that is glamorous. All of it is the stuff that gets skipped when somebody builds a site in a weekend and calls it done. Half-assed structure gets half-assed citations.
Yesterday I went through every live client site I host and checked the AI-search basics: structured data, an llms.txt file, a robots.txt that actually welcomes AI crawlers, and a sitemap that exists somewhere other than in theory. About ten sites needed work. Now they have it, and I checked each one live afterward.
The fun find? A few sites still had their canonical tags pointing at an old preview address instead of the real domain. That’s like forwarding your mail to the house you moved out of and wondering why nobody writes. Small fix, big difference in how search engines and AI tools understand which site is the real one.
Does that guarantee any of those businesses gets recommended by ChatGPT tomorrow? No, and anybody who promises that is selling something. It just stops the site from making the robot’s job harder than it needs to be. That part I can control, so I do. It’s baked into my AI search optimization work and my monthly care plans.
Yes. You sign in, verify your site, and you can import it straight from Google Search Console if you already use that. It takes about ten minutes, most of which is finding the right login.
Not yet. Right now it shows citations and queries, not clicks, so it tells you where you’re mentioned rather than who walked in the door. Pair it with your analytics to get the full picture.
No. Nothing does. What it does is show you which subjects AI already connects to your site and where competitors are getting cited instead, so you know exactly which pages to build or fix next.
If you’ve never looked at what AI thinks your business is, now’s a damn good time to start, before a chatbot decides for you. I’m happy to dig through your site, see which topics you own and which ones you’re leaving on the table, and fix the boring parts that make the difference. Start here and tell me what you do. I’ll tell you if the robots agree.
