Bing's Topic Portfolio beta: find AI-citation opportunities without guessing
A first look at Bing's Topic Portfolio beta in Webmaster Tools: how to read Topic Size, Sites Cited and Queries, turn opportunity queries into a GEO content plan, and stay honest about what it can and can't promise.
If you are trying to improve how often your site shows up in AI answers, you have probably hit the same wall: you can see where you are cited, but not what to do about it. Bing's Topic Portfolio beta — new in Bing Webmaster Tools, built on AI Performance — is a step toward filling that gap. It moves from per-query citation labels to a portfolio of topics relevant to your site, and it surfaces the queries where your site is being missed.
This guide is a plain-spoken walkthrough: what changed, how to use the opportunity queries for GEO content planning, and — because it is still a beta — how much trust to put in it.
If you are new to the wider topic, start with our generative engine optimization guide for the full playbook.
What changed: from citations on queries to a portfolio of topics
Bing's AI Performance in Webmaster Tools already showed you where your site appears in AI-generated answers, with topic and intent labels attached to individual queries. Topic Portfolio builds on exactly those labels — so if you have been using AI Performance, this is the same data, re-organised around topics instead of single queries.
The shift matters because it changes the question you can ask. Instead of "which of my pages got a citation?" you can now ask "which broader subject areas is my site actually known for in AI answers?" Rather than a long list of individual queries, you get clusters — a portfolio of the topics associated with your citation activity.
Bing describes three capabilities in the beta:
- See your topics together — a portfolio of topics relevant to your site, not just labels on individual queries.
- Add context to content planning — explore Topic Size, Sites Cited, and Queries to understand the scale and breadth of citation activity across each topic.
- Discover queries worth exploring — find grounding queries relevant to your site where other sites are cited but yours isn't. Open a topic to review these opportunities alongside queries already driving your citations.
That third piece — the opportunity queries — is the actionable core for anyone doing GEO.
How to use it for GEO content planning
The honest framing matters: Topic Portfolio is a planning signal, not a guarantee. It shows you where the gaps are. It is on you to decide which topics are worth building, and to build them well enough that an AI answer actually cites your site.
Here is a practical content-triage workflow using the metrics Bing exposes — Topic Size, Sites Cited, and Queries:
1. List your topics and check the scale
Open the topic portfolio and look at your topics together. For each, note Topic Size (how big the subject area is) and Sites Cited (how many sites Bing's answers are drawing from). A topic that is large and cites many sites is an active, contested space — that is where the competition for a citation is real, and where a strong, data-backed page can stand out.
2. Separate "topics we own" from "topics we're missing"
Split the list two ways:
- Topics where your site already drives citations. You are established here — the job is maintenance, staying fresh, and covering gaps that turn individual citations into a broader topical footprint.
- Topics where other sites are cited and yours isn't. These are your opportunity queries. This is where Topic Portfolio earns its keep: instead of guessing which queries Bing's answers are grounding on, you can see the actual grounding queries your site is missing.
3. Prioritise with intent
Not every opportunity query is worth chasing. Weight them by:
- Topic Size — a large topic with lots of cited sites is a space worth entering if it maps to a business goal; a small or tangential topic may not be.
- Relevance to what you sell — an opportunity query is only valuable if answering it well helps the people you serve find you.
- Sites Cited vs. your gaps — where the cited sites are weak on data or specifics, there is room for a genuinely better answer.
4. Build or strengthen, then measure
This is where GEO actually happens: write or expand content that genuinely answers those grounding queries — with specific, verifiable, structured information an AI answer can pull from. Then treat the topic as a hypothesis and measure whether anything changed. If the citations don't move, that is information too. Every change gets a verdict, misses included.
The point of GEO is not a single number — it is the slow, compounding build of topical authority, tracked over weeks, in a channel where citation changes materialise slowly.
Honest on limitations
Topic Portfolio is a beta, and it is evolving. Three limitations are worth holding in mind so you don't over-read the data:
- The labels are AI-generated and may occasionally be inaccurate. A topic grouping or an intent label can be wrong. Treat the portfolio as directional, and sanity-check any individual topic before you bet a content sprint on it.
- Opportunity queries don't guarantee citations. Just because other sites are cited on a query and yours isn't does not mean publishing will get you cited. It is a planning signal, not a promise.
- Bing is one AI answer ecosystem. Bing AI Performance reports on citations in Bing's AI answers, but it is not the whole AI-search picture. ChatGPT, Perplexity and Google's AI Overviews are separate systems with separate citation behaviour.
The same honesty applies to the measurement side: Bing Webmaster data can lag, and AI-citation changes materialize over weeks, not days. If you are tracking your own progress, treat each observation as directional until you have enough history to compare like-for-like.
Tying it to your wider AI-search citation strategy
Topic Portfolio is a useful input to a GEO workflow, but it is one view of one ecosystem. A durable AI-citation strategy pairs that input with:
- A baseline of where you're cited across engines. Bing tells you about the Bing/Copilot world; the other engines need their own checks.
- A way to connect search data to your own analysis. If you want to analyse this data in your own workflow, our Search Console MCP integration and the SEO MCP server guide show how to bring that data into an agent workflow.
- A measurement habit. Publish, then check whether the signal moved — and say so plainly when it doesn't.
If you want a broader picture of how your brand shows up across the AI answer engines, Cogny's AI Visibility Index asks marketing-tool questions across ChatGPT, Claude, Gemini and Perplexity and tracks which brands get mentioned — a complementary, cross-engine view of citation-style visibility.
The bottom line
Bing's Topic Portfolio beta is a genuinely useful addition for anyone doing GEO: it turns scattered citation labels into a topic-level view, and it surfaces the opportunity queries where your site is absent but others are cited. That is exactly the kind of planning signal that lets you spend content effort where it can move the needle, instead of guessing.
Just keep the frame honest: it is a beta, the labels can be wrong, and an opportunity query is a lead, not a guarantee. Use it to prioritise what to build and strengthen — then measure what actually changed.
Cogny is built for that knowing → doing → measuring loop. Your people decide; every change gets a verdict, misses included. See the plans on our pricing page, and ship a change to see it measured.
Start free — no card required.