Bing's Topics Portfolio shows who else gets cited. Which AI topics should you defend, contest or skip?

KEY TAKEAWAYS

Topic Size counts citations, not searches. It ranks topics within Bing and nothing more.
Sites Cited is a count, not a difficulty score. Divide Topic Size by it to see how concentrated a topic is.
The portfolio only lists topics where you already appear. New topics need prompt testing outside the tool.
Use the topic list as hypotheses for ChatGPT and Google. Measured overlap between engines is low.

Until this week, Bing's AI citation report mostly showed you yourself. Citation Share gave you a slice of each query, but not how big a whole topic was or how many sites were in it. Bing's Topics Portfolio adds those two numbers. Read one column at a time, they mislead. Read together, they tell you where to spend effort.

What did Bing add on 13 October?

A beta tab called Topics Portfolio, inside the AI Performance report in Bing Webmaster Tools, next to Citations & Pages. It groups the grounding queries where your site is cited into broad topics. For each topic it shows three approximate figures:

Topic Size. Total citations across all sites for the topic in the period you choose.
Sites Cited. How many sites appear for the topic's queries.
Queries. How many distinct grounding queries the topic holds.

A topic's detail page adds your own citations in that topic, a table of queries where you are already cited, and a table of opportunity queries where other sites are cited and you are not. Opportunity volumes are shown as ranges. You can pick a window of one, two or three months.

Microsoft lists the limits itself in its announcement post. Topic labels are generated by AI and can be wrong. Topics with too little presence or activity are hidden. A rival's branded query can show up as your opportunity. All figures are approximate. Microsoft's post advises treating them as "directional signals for planning."

This is the third layer on a report that opened in public preview on 10 February 2026. That release counted citations across Microsoft Copilot, AI summaries in Bing and select partner integrations. In June, Bing added Intents, Topics, Citation Share and Compare. Citation Share is your citations for one grounding query divided by all citations for that query. What is new now is the denominator at topic level, plus a list of what you are missing.

Why can't you read Topic Size as demand?

Because it counts citations, not searches. PPC Land pointed this out in its coverage. A topic can look large because many people ask about it, or because each answer in it cites many sources. The column cannot separate the two.

That matters most when you compare across engines. Profound's data put Copilot at about 2.5 domains per response, against about 7.7 for Google AI Overviews and 5.0 for ChatGPT. Profound does not state the sample behind that table, so treat it as a rough guide. The direction is enough. A citation total from one engine does not translate into another engine's terms.

Inside Bing, Topic Size is still useful. It ranks your topics against each other. That is all it should be asked to do.

Is Sites Cited a difficulty score?

Some guides read it that way. One suggests a double-digit count means a fight and a handful means an opening, which the author labels as a personal rule of thumb. The trouble is that a count says nothing about concentration. Forty sites can share a topic evenly, or two of them can take most of the citations while 38 get one each.

You can get closer with one division. Topic Size divided by Sites Cited gives the average citations per cited site. A high average means a few sources are cited again and again. A low average means the topic is spread thin and no one owns it.

Two ratios that turn the tab into a decision

Neither ratio is a Microsoft metric. Both are our derivation from the figures the tab shows.

Topic share = your citations in the topic / Topic Size. This is the topic-level version of June's per-query Citation Share.
Standing = your citations in the topic / (Topic Size / Sites Cited). This compares you with the average cited site. Above 1, you out-cite the typical source. Below 1, you trail it.

Share tells you how much of the topic you hold. Standing tells you whether you are a regular or a guest. Add a third question that no tool can answer: does the topic sell what you sell?

A worked example

This example is hypothetical. A payroll software company opens Topics Portfolio with a three-month window and sees three topics.

Topic	Topic Size	Sites Cited	Your citations	Topic share	Avg per site	Standing
Payroll compliance	12,000	300	900	7.5%	40	22.5
HR software comparisons	20,000	80	200	1.0%	250	0.8
Small business taxes	30,000	1,500	60	0.2%	20	3.0

The ratios are our calculation on invented numbers.

Payroll compliance is a strong position. The company is cited more than 22 times as often as the average source. The job is to defend it.

HR software comparisons looks like an opening by the count rule, with only 80 sites. The ratios say otherwise. The average cited site collects 250 citations, so a small group of sources is cited heavily, and the company sits just below average. New pages alone rarely move that. The useful step is to run the opportunity queries in Copilot and see which sources keep appearing. If they are review sites or publishers, getting covered by them is a press task, not a page task.

Small business taxes is the biggest topic and the most fragmented. A share of 0.2% looks poor, but standing of 3.0 means the company already beats the typical source in a field no one owns. If the topic converts for payroll buyers, clear answer pages can win ground here. If it does not, size is not a reason to chase it.

What should you do with each topic?

Pattern	What it means	Suggested call
Standing well above 1, topic fits the business	You are a regular source	Defend. Keep cited pages current and cover the opportunity queries inside the topic.
High average per site, your standing below 1	A few sources hold the topic	Contest through the sources that get cited, not only your own pages.
Low average per site, topic fits	No one owns it	Test. Publish direct answer pages for the opportunity queries.
Any pattern, topic does not fit	Visibility without buyers	Skip, however large it is.
Opportunity query names a rival brand	Not a real gap	Drop it unless you sell that product.

The cut-offs are judgement. As a rule of thumb of ours, treat standing above 2 as solid and below 0.5 as a guest appearance, then check each topic against what you sell.

What will the tab never show you?

Topics you are absent from. Microsoft describes the list as topics where your site has a meaningful presence. Opportunity queries sit inside those topics. A subject where you have no citations at all will not appear, so the tab cannot tell you that a whole category is passing you by. That needs prompt testing outside the tool, sized as in how to measure AI visibility when the answers keep changing.

It also shows no clicks. Whole Whale reported 48,000 citations across 34 pages over about three months, and 14 clicks from Copilot. That is one site, so it proves nothing general. It is a reminder that a citation is exposure, not a visit. Microsoft Clarity is the place to pair the two, since its Citations dashboard reports AI referral traffic beside share of authority.

One more trap. Many sites saw AI Performance numbers jump from 1 June. Microsoft's Krishna Madhavan described this as routine backfilling, according to PinMeTo. Any Compare view that spans that date needs a note.

Does Bing data tell you about ChatGPT and Google?

Partly, and less than it seems. The evidence points in two directions.

For a link: OpenAI's help page says ChatGPT search sometimes partners with other search providers and names Microsoft among them.

Against a close link: in a test published by Search Engine Land on 30 September, a page that only Googlebot could crawl was quoted by ChatGPT Plus. The test was one invented term on one page, and the free logged-out ChatGPT did not find it. Ahrefs, across 15,000 long-tail queries collected in July 2025, found Copilot's citations lined up most with Bing's top 10, while ChatGPT's matched Bing's top 10 only 8.1% of the time. Profound measured cited-domain overlap between engines from 6% for AI Overviews against Copilot to 16.4% for Perplexity against AI Overviews.

The weight of evidence supports a narrow reading. The figures describe Microsoft's surfaces. Bing's own reach is not trivial: StatCounter put it at 9.84% of US search in September 2026, against 85.23% for Google. But what transfers to other engines is the topic map, not the numbers. Use the list of topics and opportunity queries as hypotheses, then check them in ChatGPT, Perplexity and Google. For the Google side, see what an AI impression in Search Console is worth.

A seven-step method

Verify the site in Bing Webmaster Tools and open AI Performance. If Topics Portfolio is missing, there is not enough data yet. Check again in a month.
Set a three-month window. Note any period that crosses 1 June 2026.
For each topic, record Topic Size, Sites Cited and your citations from the detail page. Work out share and standing.
Mark each topic as fits or does not fit the business. Drop rival brand queries from the opportunity lists.
Sort topics with the decision table into defend, contest, test and skip.
For contest topics, run the opportunity queries in Copilot and log which sources recur. For test topics, write answer pages Microsoft's own guidance favours: specific headings, a short answer near the top, tables and lists, nothing important hidden in tabs or PDFs.
Run the same topics as prompts in ChatGPT, Perplexity and Google. Act on a topic in those engines only when the prompts confirm the gap.

Where this sits in digital authority

Topics Portfolio reads one pillar, AI visibility, on one family of engines. Its most useful output points at another pillar. When the same few sources hold a topic, the fix is press coverage in those sources, which is how the engines choose whom to name, as covered in how AI engines choose which brands to name. If you want to see which sources shape the answers in your category across engines, the waitlist is open.

Questions

What is Topics Portfolio in Bing Webmaster Tools?

A beta tab in the AI Performance report, added on 13 October 2026. It groups the grounding queries where your site is cited into topics and shows each topic's approximate total citations, number of cited sites and number of queries. It also lists opportunity queries where other sites are cited and yours is not.

Is Topic Size the same as search volume?

No. It counts citations across all sites, not searches or users. A topic can look large because answers in it cite many sources. Use it to rank your topics against each other inside Bing, not to forecast traffic or to compare with other engines.

Does Bing AI citation data tell me how I do in ChatGPT or Google?

Only loosely. Microsoft scopes the report to Copilot, AI summaries in Bing and select partners. Studies find low overlap between the sources different AI engines cite. Treat the topic list as a set of hypotheses and check other engines with your own prompt runs.

Why is the Topics Portfolio tab missing for my site?

Microsoft shows it only when there is sufficient data, and hides topics with too little site presence or ecosystem activity. No threshold is published. Smaller sites may see a short list or no tab at all, which does not mean they are never cited.

Sources

Bing Webmaster Blog, Introducing Topics Portfolio (beta) in AI Performance blogs.bing.com
Search Engine Land, Bing Webmaster Tools AI performance report new Topics Portfolio searchengineland.com
PPC Land, Bing shows sites the AI search queries where rivals get cited and they don't ppc.land
Relevant Audience, Bing Topics Portfolio Beta: AI Citation Gaps by Topic relevantaudience.com
Bing Webmaster Blog, Introducing AI Performance in Bing Webmaster Tools Public Preview blogs.bing.com
Bing Search Blog, New AI Visibility Insights in Bing Webmaster Tools: Intents, Topics, Citation Share, Compare blogs.bing.com
Search Engine Journal, Bing AI Dashboard Maps Grounding Queries To Cited Pages searchenginejournal.com
PinMeTo, Bing's AI citation surge was backfill, not new visibility pinmeto.com
Microsoft Clarity, Citations in Microsoft Clarity Now Generally Available clarity.microsoft.com
Microsoft Advertising, Optimizing Your Content for Inclusion in AI Search Answers about.ads.microsoft.com
Ahrefs, Only 12% of AI Cited URLs Rank in Google's Top 10 for the Original Prompt ahrefs.com
Profound, Answer Engine Citation Overlap Strategy tryprofound.com
OpenAI Help Center, ChatGPT search help.openai.com
Search Engine Land, ChatGPT is using Google Search: We tested it searchengineland.com
Whole Whale, Bing Webmaster analytics for AI search is sobering wholewhale.com
StatCounter, Search Engine Market Share United States of America gs.statcounter.com
Ecommerce Paradise, Bing Shows Which AI Queries Cite Rivals, Not You ecommerceparadise.com

Esh

Founder, Ethos Engine. Builds the Digital Authority Score and runs the placement marketplace. esh@ethosengine.io

NEXT STEP

See where your brand stands.

Your Digital Authority Score, who AI recommends for your target questions, and the placements that close each gap. Your first report is free.

Get your free report →