How to check if Copilot recommends your product

To check whether Microsoft Copilot recommends your product, ask the questions a buyer asks while deciding — "best X for Y", "X alternatives", "X vs Y" — and look for your name in the answer. Then do the part that decides the outcome: read the sentence around the mention, and read the Bing sources underneath it. Being listed is not being recommended, and on a search-grounded assistant the difference is usually traceable to one cited page.

The manual way (and why it stops working)

Open Copilot at copilot.microsoft.com and ask like a buyer, not like a marketer. "Best help desk software for a 30-person support team." "Acme alternatives." "Is Acme worth it for a small support team?" Find your name, read the sentence it sits in — that sentence is the actual result — then glance at the numbered citations, because one of them is the Bing page the sentence came from. Five or six phrasings, ten minutes, no cost, and you'll know roughly where you stand today.

Then the second pass arrives, and the cracks show fast.

• Buying intent fans out. "Best", "cheapest", "alternatives to", "X vs Y", and one variant per segment and team size — each is a separate answer with its own shortlist and its own framing of you. Covering that honestly is a morning, and it's a morning you owe again next month.
• Copilot re-searches Bing on every ask. It retrieves fresh each time, so Tuesday's flattering sentence and Thursday's lukewarm one can both be real. One reading can't separate a genuine slip from ordinary churn.
• You will read yourself into a conclusion. Skimming prose for your own name, "also worth a look" files itself under win with remarkable ease. Extracting the sentence and judging it deliberately is a different exercise from scanning for a highlight.
• The Bing source behind the sentence gets lost. The citation that produced an unflattering claim is sitting right there in the numbered list, and it is the single most useful thing on the page — but only if you note which one it was before you close the tab.
• Nothing is stored and nothing is scheduled. In two months you'll want to know whether the framing improved after you rewrote the pricing page. Chat history is not a dataset, and it can't page you the week a competitor takes your slot.

Do it with one API call

One POST to AgentGEO returns the raw record for a query: the full answer text and every Bing source cited, structured. No headless browser, no selectors to keep alive — the managed-scraper engine is maintained for you, across all six engines.

curl -X POST https://api.agentgeo.org/v1/fetches \
  -H "Authorization: Bearer ag_live_..." \
  -H "Content-Type: application/json" \
  -d '{"query": "Acme alternatives for help desk software", "surfaces": ["copilot"], "country": "US"}'

What comes back is the whole record: the answer in full, plus every Bing page Copilot cited, in rank order. The signal you want lives in two fields — the sentence in answerText that names you, and the sources[] entry it was most likely written from:

{
  "id": "run_8f3ac21b",
  "query": "best help desk software for a 30-person support team",
  "surfaces": ["copilot"],
  "status": "completed",
  "recordsDelivered": 1,
  "creditsCharged": 1,
  "answers": [
    {
      "surfaceKey": "copilot",
      "status": "delivered",
      "answerText": "For a 30-person support team, a few tools come up repeatedly. Globex is the usual first pick for its shared inbox and reporting depth. Acme is a lighter option that teams choose for fast setup, though its analytics are thinner than Globex's. Initech rounds out the list for companies already inside the Microsoft stack.",
      "sources": [
        { "title": "Globex Help Desk - Features", "url": "https://globex.example.com/help-desk", "position": 1 },
        { "title": "Best help desk software in 2026", "url": "https://example.com/roundups/help-desk", "position": 2 },
        { "title": "Acme vs Globex, compared", "url": "https://example.com/compare/acme-globex", "position": 3 }
      ],
      "fetchedAt": "2026-07-22T09:14:22Z",
      "latencyMs": 8231
    }
  ]
}

Now write the check that matters. Here Acme is named as "a lighter option" with a caveat about analytics — a mention, not a clean win — and the comparison page at position 3 is the likeliest source of that framing. Instead of testing whether the brand is merely present, split the answer into sentences, keep the ones your brand appears in, and print the sources alongside them so the framing and the Bing page behind it stay attached:

import re
import requests
from urllib.parse import urlparse

KEY = "ag_live_your_key_here"
BRAND = "Acme"

# Buying intent - the questions asked immediately before a decision.
QUERIES = [
    "best help desk software for a 30-person support team",
    "Acme alternatives for help desk software",
    "Acme vs Globex for customer support",
    "cheapest help desk tool with a shared inbox",
]

MENTION = re.compile(rf"\b{re.escape(BRAND)}\b", re.I)


def sentences(text):
    # Split prose and bullet lines alike into quotable units.
    parts = re.split(r"(?<=[.!?])\s+|\n+", text)
    return [p.strip(" -*\t") for p in parts if p.strip()]


def host(url):
    h = urlparse(url).netloc.lower()
    return h[4:] if h.startswith("www.") else h


for query in QUERIES:
    resp = requests.post(
        "https://api.agentgeo.org/v1/fetches",
        # Copilot honours country; it does not take language or web_search.
        json={"query": query, "surfaces": ["copilot"], "country": "US"},
        headers={"Authorization": f"Bearer {KEY}"},
        timeout=200,  # the API waits up to 180s for live surfaces
    )
    resp.raise_for_status()

    for answer in resp.json()["answers"]:
        if answer.get("status") == "failed":
            # A slow scrape fails at zero credits - see the gotchas to redeem it.
            print(f"{query!r}: no record this run")
            continue

        text = answer.get("answerText") or ""
        hits = [s for s in sentences(text) if MENTION.search(s)]

        print(f"\n{query!r}")
        for line in hits:
            print(f"  says: {line}")
        if not hits:
            print("  not named at all")

        # Copilot searched Bing before it wrote. These are the pages it read.
        sources = sorted(answer.get("sources") or [], key=lambda s: s["position"])
        print(f"  written from ({len(sources)} source(s)):")
        for src in sources:
            print(f"    {src['position']:>2}. {host(src['url'])} - {src['url']}")

The output isn't a score, it's evidence: for each buying query, the exact sentences Copilot wrote about you and the ranked list of Bing pages it wrote them from. Four queries against one surface is four delivered records — four credits, with failed records costing nothing. An ag_test_ key returns labelled demo records at zero credits while you shape the parser.

Read the two columns together and the fix usually names itself:

Let your agent do it

This job suits an agent better than a script, because the interesting step is the second one: fetch the answer, then go read the Bing page that produced it. One command connects the server:

claude mcp add agentgeo -- npx -y agentgeo-mcp --key ag_live_...

That's the install: agentgeo-mcp over stdio, zero npm dependencies, Node.js 18+, and a single tool called fetch_raw_answers. From there you describe the task rather than implement it:

Fetch Microsoft Copilot's answer for "best help desk software for a 30-person support team". Quote the sentence that mentions Acme, then open the Bing source it most likely came from and tell me whether that page actually says it.

Notice where the judgment sits. AgentGEO hands back the record and stops; your agent — which knows your positioning, your pricing and what you'd call a fair description — decides what counts as a recommendation and whether the cited page needs correcting or out-writing. That split is the whole argument for a data layer rather than a dashboard, and you can drive it from Python or curl just as easily.

Common problems

Five things that will make this check lie to you, in rough order of how often they do:

• Substring matching over-reports. A bare "acme" in text also fires on "AcmeBank", "Acme Freight" and the phrase "the acme of support". Use a word-boundary regex and keep an explicit ignore list for the namesakes you already know live in your category.
• Make sure you're reading the right Copilot. This is consumer Microsoft Copilot at copilot.microsoft.com — the one grounded in a live Bing search — not GitHub Copilot and not the M365 Copilot inside Word and Teams. The copilot surface targets the consumer assistant; the enterprise ones answer over private tenant data and won't match what a buyer sees.
• One run is not a verdict. Copilot re-searches Bing on every ask, so a mention that disappears once has probably not disappeared. Re-fetch two or three times before you tell anyone the framing got worse.
• The Bing citation is a lead, not proof. The model paraphrases and merges several results, so the sentence about you may be assembled from two of them rather than lifted from one. Treat the source list as where to start reading, not as a citation in the academic sense.
• Slow scrapes come back failed. Requests wait up to 180 seconds; a scrape that runs longer returns status: "failed" at zero credits with a providerFields.snapshot_id. Retry with that id and the same single copilot surface to redeem the finished answer instead of paying to re-scrape it.

Summary

For a one-off answer to "does Copilot recommend us?", ask it yourself and read the sentence. When the question has to be asked every month, across the queries that actually drive revenue, and answered with quotes instead of impressions — fetch the records, keep the sentences, and keep the Bing sources next to them. The sentence tells you how you're described; the citation tells you which page to go and fix.