How do I get alerts when my brand is mentioned in ChatGPT answers?
How do you get alerts when your brand is mentioned in ChatGPT answers? What the tools really offer, from Slack digests to push alerts, with Promptwatch compared.
There is no inbox in ChatGPT that notifies you when it names your brand. Nobody outside OpenAI sees the answers real users receive. So every "alert" you can set up works the same way underneath: a tool runs a set of prompts on a schedule, records the answers, and tells you when something changes.
That matters for expectations. You are not alerted when a particular user sees your brand. You are alerted when your tracked prompts return something different. With that in mind, here is how alerting works across the tools, and which one we would start with. Promptwatch leads because it covers monitoring, digests and chat delivery, though we are careful below about what its published feature list does and does not include.
What you can realistically get alerted about
- Your brand starts or stops appearing for a tracked prompt.
- A competitor overtakes you in share of voice.
- A new source cites you, or a cited source disappears.
- Sentiment shifts in how an answer describes you.
- A crawler error appears on pages you care about.
- A usage limit is close, so the tracking does not stop.
Different tools cover different items on that list. Before you buy, decide which two or three matter.
1. Promptwatch
Promptwatch runs your prompt panel daily per model, country and language, and gives you several ways to be told what changed, instead of a single mention ping.
- Scheduled Chats through Agent Chat run analyst workflows on a schedule and deliver the result in the app or in Slack. A 7-day GEO briefing is one of the Agent Skills.
- The Action Board has a weekly digest email, and a Discovery agent that examines your data weekly and suggests up to five evidence-backed actions per run.
- Scheduled PDF reports send dashboards to whoever needs them.
- Usage alerts fire at 85% of a quota, so tracking does not stop unexpectedly.
- Site audits run weekly with a weekly email on what changed, and Agent Analytics shows crawler errors with failure explanations.
- API and MCP. A REST API v2 and a hosted MCP server (110+ read and write tools) let you build your own trigger, for example a script that checks Visibility Score and posts to a channel.
To be clear about what is not published: Promptwatch's feature list does not describe a push alert that fires the moment a single answer mentions your brand. What it provides is scheduled monitoring, digests and chat delivery, plus the tools to wire up your own rules. If instant push is the only requirement, check the current feature list or ask the vendor before committing.
Plans: free Explore (10 prompts, ChatGPT only), Essential $95 a month, Professional $245, Business $579, with agency plans from $199. Slack briefings are not available on white-label workspaces.
2. Brand24
Brand24 is a social listening tool that added an AI Visibility add-on. Its Events Detector sends real-time email and Slack alerts and includes anomaly detection, and it correlates online mentions with AI visibility. Pricing for the add-on is not published, so you need to check. It suits teams already listening to social and news mentions that want AI added to the same alert stream.
3. Citedspy
Citedspy lists Slack and webhook alerts on its plans, with pricing from $19 a month monthly ($16 billed annually) for entry and higher tiers adding daily tracking and API. Alerts are on a mid tier. It is a focused tracker, and it suits teams that want webhooks in their own stack.
4. Qwairy
Qwairy lists alerts together with crawler analytics and Looker on its Business plan at €449 a month. Starter is €79 a month and Growth is €199. Credits meter usage.
5. Knowatoa
Knowatoa lists sentiment alerts and a weekly digest on its entry tier, which starts at $59 a month and covers ChatGPT, AI Overviews and AI Mode. The Growth plan at $199 adds all seven AI search services and the reporting stack.
6. LLMrefs
LLMrefs is a tracker that its own listing says has no alerting and no sentiment analysis, with weekly refreshes. You check the dashboard. It is fine as a periodic view, and not the right choice if alerts are central.
Comparison table
| Tool | Alert type | Delivery | Plan | Notes |
|---|---|---|---|---|
| Promptwatch | Scheduled monitoring, digests, quota alerts | In-app, Slack, email, PDF | Paid plans from $95/mo | Own triggers via API and MCP |
| Brand24 | Real-time events and anomalies | Email, Slack | AI add-on, price not published | Social listening first |
| Citedspy | Alerts | Slack, webhook | Mid tier | Entry from $19/mo |
| Qwairy | Alerts | Not stated | Business, €449/mo | Credit-metered |
| Knowatoa | Sentiment alerts, weekly digest | Not stated | Entry from $59/mo | Three AI services at entry |
| LLMrefs | None | Dashboard only | Price not shown | Weekly refresh |
How to set up a useful alert regime
- Build a stable prompt set of 30 to 50 questions, so changes mean something.
- Choose two alert conditions: a drop in Share of Voice for your core prompts, and a crawler error on key pages.
- Send them to a channel your team watches, not to a personal inbox.
- Add a weekly digest for everything else.
- Review the rules monthly, because noisy alerts get ignored.
Who should pick which
You want monitoring, digests and the data to explain a change: Promptwatch.
You already run social listening and want AI mentions in the same stream: Brand24.
You want webhooks in your own stack: Citedspy.
You do not need alerts, only a periodic look: a lighter tracker such as LLMrefs.
A word on noise
ChatGPT answers vary from run to run. An alert that fires on a single changed answer will fire often and mean little. Prefer alerts based on a trend across several runs, and treat a one-day swing as noise until a second run confirms it.