> ## Documentation Index
> Fetch the complete documentation index at: https://docs.openpulse.cloud/llms.txt
> Use this file to discover all available pages before exploring further.

# Listeners: what you are watching for

> A listener is one standing question, asked on a schedule. You pick an objective, and the objective sets the sources, the data window and the starting score bar.

A **listener** is one standing question, asked on a schedule. You pick an objective, and the objective sets the sources, the data window and the starting score bar.

## Four objectives

| Objective                                   | What it finds                                                                                                                                                  | Sources                                                                          | Window  |
| ------------------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------- | -------------------------------------------------------------------------------- | ------- |
| **Find customers**                          | People describing the problem you solve, asking for recommendations, or complaining about the workaround they use instead                                      | Reddit · Hacker News · X · LinkedIn · Facebook                                   | 30 days |
| **Situations that mean somebody needs you** | A company hires three support agents, signs a lease, raises a round, or says their vendor is slow — read against what *you* sell rather than against a keyword | Reddit · Hacker News · X · LinkedIn · Ashby · Greenhouse · Lever · LinkedIn Jobs | 30 days |
| **Competitor dissatisfaction**              | People publicly fed up with a named rival: switching, frustrated, comparing                                                                                    | Reddit posts *and* comments · Hacker News · YouTube · G2 · Trustpilot · X        | 30 days |
| **Local prospects**                         | Google reviews for the businesses you sell to, themed and scored                                                                                               | Google Reviews                                                                   | 90 days |

The first two run on the same engine and differ in what they are looking for; the last two each have their own vocabulary and their own roll-up view — see [Competitor intelligence](/guides/competitor-intelligence) and [Local prospects](/guides/local-prospects).

## Creating one starts with your website, not a keyword box

Paste a URL. Openpulse crawls the site and proposes the whole plan: keywords, exclusions, per-source search queries, and a profile specific to the objective (target roles, rival names and pain vocabulary, or place categories).

**All of it is editable before anything runs**, and editable afterwards too. The plan is not frozen at creation — you can change keywords, exclusions and queries on a running listener without losing its signals.

This same crawl-and-plan step is available programmatically as `POST /v1/listeners/analyse-website` — see the [API reference](/api-reference) — and to an AI assistant as the `listeners:analyse` MCP scope, described in [MCP & agents](/mcp/tools).

## Know how much of a post each source returns

A source that cannot hand over the post body leaves the classifier judging a search snippet, which is the single biggest cause of a bad label. Every source is labelled in the wizard so this is never a surprise:

| Label           | Sources                     | What it means                                                                      |
| --------------- | --------------------------- | ---------------------------------------------------------------------------------- |
| **full text**   | Reddit, Hacker News         | The whole post, its author, and its engagement counts                              |
| **snippet**     | LinkedIn, X, the job boards | A search-engine snippet: enough to spot a topic, not always enough to judge intent |
| **best effort** | Facebook                    | No provider sells keyword post search here. Expect little.                         |

## Schedules

Four cadences: **hourly**, **every 6 hours**, **daily** and **weekly**. Which ones you can pick depends on your plan — cadence is the main difference between Go and Pro, because every run costs real retrieval and model spend. See [openpulse.cloud/pricing](https://openpulse.cloud/pricing) for which cadences each plan includes.

You can stop a listener's schedule and start it again later. Stopping removes it from the queue entirely rather than firing jobs that get discarded on arrival; a run already in flight finishes its pass. The listener panel shows the live schedule state, including the next fire time.

## The listener panel

A side panel with five tabs, reached from a listener card or from its signals:

* **Overview** — what it is watching, its score bar, its status
* **Search plan** — keywords, exclusions, per-source queries, all editable
* **Pipeline** — the rule that puts strong signals straight onto the deal board
* **Emails** — who gets this listener's run summaries
* **Activity** — recent runs, with their step logs

*Run now*, *Stop schedule* and *Delete* stay on screen while you scroll, so you are never acting on a listener whose name has scrolled away.


## Related topics

- [What Openpulse does](/index.md)
- [Runs and signals](/guides/runs-and-signals.md)
- [MCP: your AI assistant reads the workspace](/mcp/overview.md)
