How do you get Copilot to recommend your brand?

Microsoft's transparency note for Copilot sets it out. It says Copilot draws on language and multi-modal models from Microsoft, OpenAI and other model developers, and that for certain conversations where a user is looking for information, it is grounded in web search results: it can search the web and use the top results to write a summary, with hyperlinked citations listed below the text so the reader can open the sites it used.

Microsoft reports those citations to site owners through Bing. Its AI Performance report, in public preview in Bing Webmaster Tools since 10 February 2026, shows how often a site's pages are cited as sources in Microsoft Copilot, in AI-generated summaries in Bing and in some partner integrations, with the key phrases the AI used when retrieving the content it cited. Copilot also began in Bing: on 15 November 2023 Microsoft said Bing Chat was becoming Copilot, with copilot.microsoft.com as its own home.

How it weighs one source against another is Microsoft's call, and the note does not set it out beyond saying the summary uses the top results. Because Microsoft reports a site's Copilot citations through Bing, the quickest levers on a brand's own side are whether Bing can crawl and index the site, and what the site says.

Access first, and Microsoft publishes three things about it:

• bingbot. Microsoft describes it as Bing's crawler, which discovers new and updated pages to add to Bing's searchable index, with the aim of keeping that index comprehensive and fresh.
• IndexNow. Microsoft says it helps keep information fresh across search and AI experiences by notifying participating search engines whenever content is added, updated or removed, rather than waiting for the next crawl.
• NOARCHIVE and NOCACHE. In September 2023, when Copilot was still called Bing Chat, Microsoft said a page tagged NOARCHIVE would not be included in its chat answers or linked from them, that a page tagged NOCACHE could appear with only its URL, title and snippet, and that either page would still appear in Bing search results.

Which crawlers to allow for the other assistants, and why a rule at your network edge can refuse one your file allows, is its own answer.

Then the pages themselves. Introducing its AI Performance report, Microsoft suggested clear headings, tables and FAQ sections, so AI systems can reference the content accurately; examples, data and cited sources, which it says help build trust when content is reused in an AI answer; regular updates, so the most current version is the one referenced; and text, images and video that describe the same products in the same way. Beyond that, the advice is ours rather than Microsoft's. Buyers usually describe a situation rather than name a product: a small team, a budget, a named use. Answer that situation directly on the page, in a passage that still makes sense read on its own.

And the facts on them. Microsoft says Copilot writes its summary from the web search results it uses, so a wrong price, date or specification on one of your pages can reach an answer. Where an answer and your page disagree, the fix can sit on either side. Which of these to start with depends on where your brand is weakest, and that order has its own answer.

You can see which of those pages Copilot leans on in your own category. Ask a buying question, open the citations listed under the answer, and note the domains: Microsoft says they are the sites used to ground the response. Where the same domains keep coming back on questions a competitor wins and you do not, those domains are where the work goes. Why a competitor comes up first has its own answer.

Those pages sit at different distances from you, and the work changes with the distance:

• A listing you supply. A retailer page, a marketplace listing, a directory profile. You can influence it directly by correcting what you supply, though the page is not yours.
• An independent review or comparison. You can give the writer accurate facts, access to people who can answer questions and, where it fits, the product to try. What they conclude is theirs.
• A forum or community thread. You can answer the question openly, under the brand's own name. The opinion belongs to the people in the thread.

None of it moves on your schedule. A page about you changes when the people who publish it decide to change it.

Introducing the updated Copilot in October 2024, Microsoft Advertising said ads would appear below Copilot's organic response, in a layout that separates organic and sponsored conversation more clearly. In March 2025 its corporate vice president wrote that organic responses are never influenced by sponsored content, which is distinctly marked and appears only when highly relevant.

The formats are moving closer to the answer. In April 2026 Microsoft described Offer Highlights, ad experiences that let a brand promote what sets a product apart within the context of a Copilot conversation, for retail in English-speaking markets, as more than a traditional ad below an AI response.

So an ad buys a marked placement beside or within the conversation, not a place in the organic answer. Count the two apart: an entry marked as sponsored is an ad, however close it sits to the recommendations, and a report that counts ads and recommendations together overstates how often a brand is recommended.

The variation is large. In Delphi's study of answers collected up to 24 August 2026, an assistant asked the same question more than once inside one measurement named the identical set of companies 21.5% of the time. A before and after built from one answer each compares two draws, and it can show a gain or a loss that nothing caused. How much evidence a change needs has its own answer.

The free check by hand is written for ChatGPT and works the same way here: it sets out which questions to ask and what to write down for each answer, and comparing a before with an after adds to what you record each time. Three things are specific to doing it in Copilot:

• Keep your own history out. Microsoft says Copilot can remember facts you ask it to and use them to make its answers more relevant, and that you can opt out of personalization at any time in its settings. Ask with it off.
• Note the links each time. Microsoft says an answer grounded in a web search lists the sites it used below the text. A page you changed can only show in an answer that searched, so note which of the answers you compare carry links.
• Allow for the clock. A rewritten page can show up only once Copilot's search reaches the new version, and Microsoft suggests IndexNow to tell search engines it changed rather than wait for the next crawl. A page about you on another site changes only once its publisher changes it.

Bing Webmaster Tools shows part of it. Its AI Performance report counts how often your pages are cited as sources in Copilot and in Bing's AI summaries, which pages, and the phrases the AI used when retrieving them. It counts your pages, so an answer that names your brand and cites somebody else's page leaves nothing in it.

The study publishes that figure with the counts behind it and how each was made. A brand's standing in Copilot is one part of its AI visibility. Some of the work itself carries across and some does not. Accurate pages and third-party coverage are there for any assistant that reads the web. Crawler access is set operator by operator: most publish their own crawler, so a rule for bingbot says nothing about the rest.

The same question has its own answer for ChatGPT, Google Overview, Gemini, Claude, Perplexity and Grok.

Delphi measures them side by side. Each time it measures a brand, it puts the same questions to ChatGPT, Gemini, Claude, Perplexity and Grok more than once, and to Google Overview and Copilot once, and keeps every answer word for word. Its score, the Delphi Index, carries a margin set by how far the assistants disagreed with each other and by which questions were asked, and the dashboard and the history page do not report a change in the Index as a change unless it is larger than the assistants' disagreement. Each assistant's share of answers naming you is shown beside the others' from one measurement to the next, so a change in Copilot alone is visible, though that figure carries no margin of its own.

What Delphi reads from Copilot is Microsoft Copilot's answer to a search on Bing: each question is searched as it is written, with no history behind it, and the sites the answer cites are kept with it. Each question is searched once in a measurement, so Copilot's share is counted across the questions rather than read from any one answer. Where Copilot gives no answer to a search, that question is left out of Copilot's figures rather than counted against the brand. So a page change can show in a Delphi reading of Copilot once the search reaches the new version. Each paid level includes one or two measurements per brand a month. It is not a daily tracker, and an extra measurement costs $49 when a before and after needs one sooner.