Perplexity and Microsoft Copilot: How to Improve Your Visibility

Key facts

• Perplexity documents PerplexityBot as the crawler that surfaces and links websites in Perplexity search results, and says it is not used to crawl content for AI foundation models.
• Perplexity-User fetches pages when a user asks a question and, according to Perplexity, generally ignores robots.txt because the fetch is user-requested.
• Microsoft documents that Microsoft Copilot and Copilot Chat, when web search is enabled, send a short generated query to the Bing search service and use the results to compose the response.
• IndexNow lets a site notify participating search engines about added, changed or deleted URLs; Bing states it does not guarantee crawling or indexing.
• No one can guarantee that Perplexity or Microsoft Copilot will mention or cite a business.

How does Perplexity find and link to websites?

Perplexity finds websites for its search results mainly through PerplexityBot, a crawler that Perplexity says is designed to surface and link websites in Perplexity search results. Perplexity recommends allowing PerplexityBot in robots.txt for a site to appear. Separately, Perplexity-User may visit a page in real time when someone asks a question, and Perplexity says it may link to that page in the response.

Perplexity is built around showing its sources, so the practical question for a business is whether its pages are in the pool Perplexity can draw from and whether those pages answer questions clearly enough to be useful. Perplexity's documentation is explicit that PerplexityBot exists to surface and link websites, and that it is not used to crawl content for AI foundation models. Allowing it is therefore a decision about search visibility, not about model training.

Perplexity-User works differently. It acts on behalf of a user who has asked a question, and Perplexity says that because a user requested the fetch, this fetcher generally ignores robots.txt rules. It is also documented as not being used for web crawling or for collecting training content. A site cannot rely on robots.txt to keep Perplexity-User out, and it does not need to do anything special to let it in.

Perplexity does not publish a ranking formula for which sources appear first, and its behaviour can change. What a business controls is access, clarity and corroboration.

What robots.txt and firewall settings does Perplexity visibility need?

Perplexity visibility needs robots.txt to allow PerplexityBot and any firewall to let verified PerplexityBot traffic through. Perplexity publishes IP ranges for PerplexityBot and Perplexity-User as JSON files and recommends matching both the user agent and those IP ranges in firewall rules. Perplexity also notes that robots.txt changes may take up to 24 hours to take effect.

Each Perplexity robots.txt setting works independently, according to its documentation. That means a site can make separate decisions for different Perplexity agents, although Perplexity-User generally does not follow robots.txt in any case.

Firewall rules based only on user-agent strings are fragile, because any client can claim to be a known bot. Rules that also check the published IP ranges are more reliable. Perplexity advises refreshing those IP ranges regularly from its JSON endpoints, because the ranges can change.

After a change, check server logs to confirm that PerplexityBot requests are receiving normal 200 responses with real content, not redirects to a login page, a cookie wall or a bot challenge.

• Allow PerplexityBot in robots.txt on public pages you want visible.
• Match firewall allow rules to both the user agent and Perplexity's published IP ranges.
• Refresh the IP ranges on a schedule rather than hard-coding them once.
• Allow up to 24 hours for robots.txt changes, then check the logs.

How does Microsoft Copilot use Bing to answer questions?

Microsoft Copilot uses Bing by turning part of a user's prompt into a short search query and sending it to the Bing search service when web search is enabled. Microsoft's documentation says the information Bing returns is then used to compose Copilot's response, and in Copilot Chat the exact web queries are shown in the linked citation section. Bing indexing is therefore the gateway to Copilot web answers.

Microsoft's documentation on web search in Microsoft Copilot and Copilot Chat describes the flow: Copilot parses the prompt, identifies terms where web information would help, and generates a search query of a few words informed by the prompt. That query, stripped of user and tenant identifiers, goes to the Bing search service. The response is grounded in what Bing returns.

The practical implication is that a business that is poorly indexed in Bing has little chance of being used in Copilot's web-grounded answers, however well it performs elsewhere. Many businesses in Saudi Arabia and Jordan focus only on Google and have never verified their site in Bing Webmaster Tools.

There is also an organisational factor. Microsoft documents that IT admins can turn web search off for Microsoft Copilot and Copilot Chat, and users can switch off a Web content toggle. In some workplaces, Copilot will not consult the web at all, which is outside any website owner's control.

What should a business do in Bing to support Copilot visibility?

To support Copilot visibility, a business should verify its site in Bing Webmaster Tools, submit an XML sitemap, check that Bing can crawl and index its important pages, and consider IndexNow to notify Bing when URLs are added, changed or deleted. Bing states that IndexNow does not guarantee that pages will be crawled or indexed, so it supports indexing rather than replacing it.

Bing Webmaster Tools is the Bing equivalent of Google Search Console: it shows how Bing sees the site, which URLs are indexed and where crawl problems exist. For a bilingual site, check that both the Arabic and English versions are indexed rather than only one language.

IndexNow works by hosting a UTF-8 key file on the site, usually at the root, and sending the changed URLs to an IndexNow endpoint along with that key. Bing's guide recommends using an existing CMS or CDN integration where one is available, and verifying in Bing Webmaster Tools that the URLs were received. It is most useful for sites whose content changes often, such as product catalogues, news or events.

The same content rules apply as elsewhere: server-rendered text, answer-first pages, clear business identity and accurate contact information.

• Verify the site in Bing Webmaster Tools.
• Submit the XML sitemap, including both language versions.
• Check indexing of priority Arabic and English pages.
• Set up IndexNow if content changes often; confirm receipt in Bing Webmaster Tools.

What content helps a business appear in both Perplexity and Copilot answers?

The content that helps in both Perplexity and Copilot is the content that helps any answer engine: pages that state the answer to a specific question in their opening lines, include concrete facts such as location, service scope and process, identify the business consistently, and are supported by accurate mentions on independent sites. Both systems show sources, so a precise, quotable passage is more useful than general marketing copy.

Perplexity and Copilot both compose an answer from retrieved sources and show links to them. A passage that answers a narrow question completely is easier to use than a page that answers several questions vaguely. Short definitions, step-by-step processes, clear eligibility criteria and comparison tables are all formats that can be quoted accurately.

Corroboration matters because both systems usually consult several pages. If directories and industry sites describe the business the same way its own site does, the answer is more likely to be accurate. If they conflict, outdated details can surface in the answer.

How can a business measure visibility in Perplexity and Copilot?

A business can measure Perplexity and Copilot visibility with a fixed set of prompts tested on a recorded date, in Arabic and English, per target country, recording brand mention, linked citation, other cited domains and prominence. Server logs show whether PerplexityBot is crawling key pages. For Bing-grounded experiences, Bing Webmaster Tools offers an AI Performance report in public preview that reports citation activity.

Microsoft describes the AI Performance report in Bing Webmaster Tools as showing how publisher content appears across Microsoft Copilot, AI-generated summaries in Bing and select partner integrations. Its metrics include total citations, average cited pages, a sample of grounding queries and page-level citation activity. Microsoft is clear that citation counts do not indicate placement, ranking or authority, and the feature is still in preview, so treat it as directional evidence.

Perplexity does not offer an equivalent site-owner report, so direct prompt testing is the main method there. Keep the prompt set stable and repeat it on a schedule; changing prompts every month makes the trend unreadable. Copilot answers can also differ between the consumer and work versions and depend on whether web search is enabled, so note which version was tested.

• Same prompts, same languages, same countries, every cycle.
• Record brand mention, linked citation, cited domains and prominence.
• Record the date and the product version tested.
• Cross-check with server logs and Bing Webmaster Tools data.

What are the limits of optimising for Perplexity and Copilot?

The limits of optimising for Perplexity and Copilot are real: neither platform publishes how it chooses which sources to cite, both change their behaviour over time, Copilot's web access can be disabled by an organisation, and answers vary between sessions. No one can guarantee that Perplexity or Microsoft Copilot will mention or cite a business. What can be improved is access, clarity, accuracy and corroboration.

Any provider promising a fixed number of Perplexity or Copilot citations is making a promise neither platform supports. There is no documented paid route into their organic answers, and no agency has privileged access to their selection processes.

A sound plan removes technical barriers first, then improves content and third-party accuracy, and measures against a dated baseline. Because these two platforms rely on different infrastructure, a site can perform well in one and poorly in the other; measuring them separately avoids false conclusions.

Frequently asked questions