Straight answers

Prompt Prospect is B2B social signal intelligence. You point it at public conversations that matter — competitors’ posts, market creators, keyword searches, or your own posts — and it identifies the people engaging there, enriches them, and evaluates each person against your ideal-customer profile.

Instead of a raw list of reactions, you get a source-backed view: who fits, why they fit, and the exact activity that brought them into view.

No. Prompt Prospect works entirely from public engagement data collected through data partners. There is nothing to install, no cookie to paste, no browser extension, and we never ask for your credentials.

Your account is never involved, so there is nothing to put at risk. Monitoring happens on our side against public data — your profile, your session, and your activity are untouched.

Two ways. Fastest: in Ideal customer, click “Scan my website” — we read your site and propose a profile you can review and save. Or describe your ideal customer in plain language and refine it with the built-in editing chat.

Your profile is one living document. Every person is evaluated against it, so sharpening it improves every match that follows.

A person’s posts (a founder, an influencer, your competitor’s CEO), a company page, a single post, a keyword search, or your own posts. Paste any public profile, company, or post URL and the type is detected automatically — or just describe who you’re looking for and let the agent find the right thing to watch.

A keyword watch finds new public posts matching a boolean query — up to six keywords with AND, OR, and NOT. You can add a plain-language filter to keep only the posts that actually mean what you mean (“complaining about their current tool, not promoting one”).

Authors of matching posts are evaluated too, with the source post kept alongside the person.

Watches are checked around the clock — new posts are picked up on an hourly cycle, and each post’s engagement is synced on a daily rhythm. In practice, a new person typically lands in your workspace, evaluated and enriched, within a day of engaging.

Instead of building a query, you ask: “Who’s complaining about legacy CRMs this week?” The agent searches public posts and shows real samples. You mark individual posts as relevant or irrelevant, then explicitly start a recurring watch when the examples fit.

Yes. Add your profile as a watch and every public liker and commenter can be captured and evaluated like any other engager. This shows who is in your audience, not just how large the reaction count is.

Every captured engager is enriched (role, company, location, headline) and then evaluated against your ideal-customer profile. The result is Strong match, Possible match, or Unlikely — with written reasoning rather than a black-box score.

The evaluation also reads what they actually did: the post they engaged with and what they said. Someone praising a competitor and someone describing a problem are different conversations, so the source and context stay attached to the result.

No. Each person is evaluated once against your current profile and remembered. If they engage again next week, they don’t consume another evaluation.

Yes — every watch has a glass-box detail page: when it was added, how many times it has been checked, the posts found, the people captured, how many have been evaluated, and the matches. If a source is being checked successfully but simply is not posting, the page says so instead of leaving you guessing.

Every first-time Strong or Possible match raises an in-app notification. For strong matches you can turn on instant email alerts. Once a day, at the hour you choose, a summary gathers everything new with the strongest matches first.

No — and that’s deliberate. Drafting is an optional downstream tool. For a person you select, Prompt Prospect can draft a connection note and email grounded in the activity they engaged with. You decide whether to use, edit, or ignore it; nothing is sent for you.

Yes — CSV or JSON, honoring whatever filters you have applied. Match decisions, written reasoning, source context, and enrichment come along.

Today: CSV/JSON export, email alerts, and an MCP server — which means your matches are queryable from Claude and other MCP-capable clients. Direct Slack, Clay, and Sheets integrations are on the roadmap.

Plans are metered on people analyzed per month — watches and teammates are unlimited on every plan. See the pricing page. During early access, onboarding is concierge: we set up your first watches with you on a call.

Public social engagement — reactions, comments, posts, and public profile data — collected through established data partners. No private data, no scraping through your account.

Not today. Enrichment covers role, company, location, and headline — email-address enrichment is on the roadmap. Every match comes with a public profile link and the public activity that surfaced the person.

If your workflow uses email, export selected people as CSV and use your existing, authorized enrichment process.

Yes. Workspaces are built in — each with its own ideal-customer profile, watches, people, and notifications. Agencies can run one workspace per client and switch between them from the same account.