How to monitor brand mentions with ChatGPT, Claude, or any AI assistant

Last updated: Nov 17, 2026

Your brand gets named in places you don't control: a review on someone's blog, a comparison page written by a rival, a news item, a Stack Overflow question about an error, a DEV tutorial. Finding them means running the same name through a search engine, a news search and a couple of developer sites, then merging what comes back. With Firecrawl Alexandria connected, an assistant does that in one pass: Firecrawl's own web and news search covers the open web, and three data providers, Stack Exchange, DEV and Particle, cover developer questions, articles and podcast conversations.

Starter prompt

Put in your brand and a time window, then send this from any chat that has Firecrawl:

Use Firecrawl Alexandria to find mentions of [brand] from the past
month. Run Firecrawl web and news search, search Stack Overflow
through Stack Exchange, list the [brand] tag on DEV, and check
Particle for podcast mentions of [brand domain]. Quote the price of
each tool before you run it.
 
For each mention, give me:
- The page title, site, date (when shown) and link
- One line on what it says about [brand]
- A label: review, comparison, question, tutorial or news
- Whether it's on [brand]'s own site, so I can leave those out
 
Only public pages. Mark any tone call as your own reading, since no
tool returns a sentiment score, and quote titles exactly.

What counts as a brand mention?

Any public page that names your company or product counts. For a weekly check, five kinds are worth separating:

  • Reviews and comparisons. Blog reviews and "X vs Y" or "alternative to" pages, which shape what buyers read before they talk to you.
  • News and announcements. Press releases, partner news and coverage that names you.
  • Developer questions. Stack Overflow posts about your product, which point at bugs and gaps in the docs.
  • Tutorials. DEV and similar articles that show people how to set your product up.
  • Podcasts. Hosts and guests naming you on a show, which a web search does not pick up.

Where to find brand mentions

SourceWhat you getThe catch
Google AlertsFree email digests when new Google Search results match your term, at the frequency you pickOnly what Google indexes and chooses to send, with no way to group or label the results
Stack Overflow and DEV searchFree question search on Stack Overflow and a browsable tag page on DEVOne site at a time, and you copy the results together yourself
Social listening toolsContinuous collection, stored history and sentiment charts. Brand24, for example, includes AI sentimentA subscription: Brand24 starts at $199 a month billed annually, capped at 3 keywords and 2K mentions
An AI assistant with Firecrawl AlexandriaWeb, news, Stack Overflow, DEV and podcast mentions in one labeled list, each with its linkUses Firecrawl credits per call, runs only when asked, and gives no sentiment score

How to monitor brand mentions with AI, step by step

  1. Give the assistant Firecrawl. The Claude, ChatGPT and MCP client sections below walk through it for each app.
  2. Name the brand and the window. If the name is also a common word, add your domain or product category so the search stays on topic, and ask it to drop your own site.
  3. Choose where to look. Web and news search covers reviews, comparisons and coverage. Stack Exchange covers developer questions, and DEV covers tutorials. The assistant finds each tool, quotes its price and runs it.
  4. Read before you respond. Every row links to the page. Check the original before you reply or cite it, and throw out matches where the name means something else.

To track brand mentions over time, keep the prompt and run it each week, or schedule the API script below.

What data can you get?

One Firecrawl tool and three Firecrawl Alexandria providers cover the main places a brand is named:

ToolWhat it returnsCredits per call
Firecrawl web and news searchRanked web and news results for your query, with title, link and a short excerpt. Search operators such as quotes and -site: work, and a time filter limits results to the past day, week or month2 per 10 results
Stack Exchange searchQuestions on Stack Overflow or another Stack Exchange site matching your text or tag, with title, tags, score, views, answer count, whether an answer was accepted, and the link. Filter by creation date5
DEV article listArticles under a tag or from an author or organization, with title, teaser, author, tags, publish date, reading time, and reaction and comment counts. No keyword search5
Particle podcast mentionsEvery line of podcast dialogue that names a company, found by its domain or slug, with the episode, show, speaker, timestamps and surrounding lines. Filter by date, show or speaker role15

On Nov 17, 2026, we ran the sweep for Firecrawl itself, leaving out firecrawl.dev and limiting results to the past month:

  • Web search: 10 results, among them a review on another scraping vendor's blog, comparison pages from competing APIs and a PR Newswire release about a Benzinga partnership, next to Firecrawl's own npm and GitHub pages, which a prompt can filter out
  • News search: 10 results, mostly videos, directories and integration listings
  • Stack Overflow: five questions naming Firecrawl, the newest from Dec 10, 2025
  • DEV: the five newest firecrawl articles, published Oct 6 to Nov 12, 2026

The sweep cost 14 Firecrawl credits. A Particle call for notion.so the same day returned podcast episodes with each line that named Notion, who said it and when in the episode.

How to monitor brand mentions in Claude

  1. Find Firecrawl in Claude's connector directory and press Connect.
  2. Sign in to Firecrawl, choose a team and approve the access request. A free account can be created during the same sign-in.
  3. Open the + menu in a chat, go to Connectors and switch Firecrawl on.

Then ask the way you'd ask a colleague:

Use Firecrawl Alexandria to search the web and news for Supabase
mentions from the past week, skip supabase.com, and sort what you
find into reviews, comparisons and news with a link for each.

How to monitor brand mentions in ChatGPT

  1. Install the Firecrawl plugin for ChatGPT.
  2. ChatGPT then shows a Firecrawl login; approve it (new users can sign up first).
  3. Open each brand-check message with @firecrawl, which routes it to the plugin.
@firecrawl Search Stack Overflow through Stack Exchange in Firecrawl
Alexandria for questions about Prisma created since Oct 1, 2026, and
list the ones with no accepted answer.

How to monitor brand mentions in Claude Code, Cursor and other AI agents

Brand checks from a terminal or IDE run on the Firecrawl MCP server, which carries the web search and every Firecrawl Alexandria provider in this sweep. Adding the server opens a Firecrawl login page; approve it and the tools load.

  • Claude Code: run claude mcp add --transport http firecrawl https://mcp.firecrawl.dev/v2/mcp-oauth; the login prompt appears under /mcp.
  • Cursor, or a client with a JSON config: paste this server block into its MCP configuration.
{
  "mcpServers": {
    "firecrawl": {
      "url": "https://mcp.firecrawl.dev/v2/mcp-oauth"
    }
  }
}

The agent gets three tools. firecrawl_search runs the web and news search and, with sources: ["alexandria"], returns providers only. firecrawl_find_tools lists a provider tool's inputs (pass includeCosts: true to see what each call costs), and firecrawl_scrape runs the Stack Exchange and DEV tools. A coding agent can also save each week's results to a file and diff them against the last run.

How to monitor brand mentions with the Firecrawl API

For a weekly report in your own stack, install the Python package (pip install firecrawl-py), call the search, the Stack Exchange tool and Particle, and schedule the script with cron:

from firecrawl import Firecrawl
 
app = Firecrawl(api_key="fc-YOUR_API_KEY")
 
def provider_data(result):
    item = result.alexandria[0]
    if item.error:  # a provider can fail without returning data
        raise RuntimeError(f"{item.provider}: {item.error.code} {item.error.message}")
    return item.data
brand = "firecrawl"
 
# Web and news mentions from the past week, 2 credits per 10 results
mentions = app.search(
    f'"{brand}" -site:firecrawl.dev',
    sources=["web", "news"],
    tbs="qdr:w",
    limit=10,
)
 
# Stack Overflow questions created since a date, 5 credits
questions = app.scrape(
    alexandria={
        "provider": "stackexchange-com",
        "capability": "questions/search",
        "options": {"q": brand, "fromdate": "2026-11-01", "pagesize": 20},
    }
)
 
for result in (mentions.web or []) + (mentions.news or []):
    print(result.title, result.url)
for q in provider_data(questions)["questions"]:
    print(q["title"], q["source_url"])
 
# Podcast episodes that name the company, found by its domain, 15 credits
podcasts = app.scrape(
    alexandria={
        "provider": "particle",
        "capability": "podcasts/mentions",
        "options": {"company_id": "firecrawl.dev", "since": "2026-11-01", "limit": 10},
    }
)
 
for row in provider_data(podcasts)["data"]:
    print(row["episode"]["podcast"]["title"], row["episode"]["title"], row["mention_count"])

All three calls run on an API key from your Firecrawl account. Keyless access covers the search but not the Stack Exchange or Particle tools, which are Firecrawl Alexandria providers. Request details for the provider tools are in the Firecrawl Alexandria docs, and the search docs list every search option.

What it costs

Looking up a tool and its price costs nothing in any of these setups. Every search you run is charged in Firecrawl credits at the rates in the data table, and the reply names each tool and its charge. Web plus news search with 10 results each, one Stack Exchange search and one DEV tag list come to about 14 Firecrawl credits, and Particle's podcast mentions add 15, so a weekly check stays small. Prices can change; Firecrawl credit bundles by plan: Firecrawl pricing.

Limits to know

  • Particle's terms come first. An admin on your Firecrawl organization accepts Particle's data terms once, before the team's first podcast mentions call.
  • No sentiment score. The tools return pages, questions and articles with their counts. Any read on tone comes from the assistant, so check the source before acting on it.
  • Search returns excerpts. Web results come back as a title and a short excerpt, and some carry no date. Ask the assistant to open the pages that matter before it summarizes them.
  • One page per call. Each search and each provider call returns one page of results, so a busy brand needs a narrower window or a follow-up call for the next page.

Related guides

Use Firecrawl in your agent

Add Firecrawl to ChatGPT, Claude, Cursor and more in one click.

Frequently Asked Questions

What is brand monitoring?

Brand monitoring means keeping an eye on where your company or product gets named in public, such as reviews, comparison pages, news stories, developer Q&A and tutorials. Catching those early lets you answer a question, correct a wrong claim or thank a writer while the page is still fresh.

Is there a free way to monitor brand mentions?

Google Alerts sends free email digests when new Google Search results match a term you set, and Stack Overflow and DEV each have a free search or tag page you can check yourself. Connecting an AI assistant to Firecrawl Alexandria runs the web, news, Stack Overflow and DEV checks from one request for a few Firecrawl credits. Prices can change.

How is this different from social listening tools?

Social listening tools are subscriptions that collect mentions around the clock, keep a history and chart sentiment. An assistant with Firecrawl Alexandria runs a check when you ask for it, pays per call and keeps nothing between runs, which suits a weekly sweep more than a live dashboard.

Can an AI assistant tell me if mentions are positive or negative?

It can read the pages it found and describe their tone, but that is the assistant's own reading. None of the search, Stack Exchange, DEV or Particle tools return a sentiment score, so open the page before you act on a mention that looks critical.

How much does it cost to monitor brand mentions with Firecrawl?

Looking up a tool's price is free. A Firecrawl web search costs 2 Firecrawl credits per 10 results, so web plus news is about 4. A Stack Exchange question search or a DEV tag list costs 5 Firecrawl credits each, which puts a full sweep at about 14. Particle's podcast mentions add 15. Prices can change.

What is Firecrawl Alexandria?

Firecrawl Alexandria is Firecrawl's library of 400+ data providers that AI assistants and agents can call directly, from public sources and licensed publishers such as Fiscal.ai, Benzinga and Apollo to Firecrawl's own Research, Developer and Government indexes. Connect Firecrawl to ChatGPT, Claude, Cursor or another agent, and it finds the right provider for your question, shows the price of each tool and returns the record from the source. It needs a Firecrawl account, signed in through the connector, plugin or MCP sign-in, or an API key; keyless connections don't include it. Finding tools is free; running one uses Firecrawl credits.

//
Get started
//
Ready to build?
Start getting web context for free and scale seamlessly as your project expands. No credit card needed.
Are you an AI agent? See setup options