Social Listening API - Audience Intelligence, Pulsar avatar

Social Listening API - Audience Intelligence, Pulsar

Pricing

from $10.00 / 1,000 mention or post returneds

Go to Apify Store
Social Listening API - Audience Intelligence, Pulsar

Social Listening API - Audience Intelligence, Pulsar

Export social listening and audience intelligence data from your own Pulsar account: TRAC searches, posts and reactions with impressions, engagement, reach and visibility, and CORE brands, profiles, post metrics and totals. Read only GraphQL queries. Bring your own token.

Pricing

from $10.00 / 1,000 mention or post returneds

Rating

0.0

(0)

Developer

Nabeel Hassan

Nabeel Hassan

Maintained by Community

Actor stats

0

Bookmarked

2

Total users

1

Monthly active users

22 days ago

Last modified

Share

Export the social listening and audience intelligence data your Pulsar account already holds into a dataset, a spreadsheet or a warehouse: TRAC searches, the posts and reactions they collected with impressions, engagement, reach and visibility, and CORE brands, profiles and post metrics. Uses your own Pulsar API token, sends read-only GraphQL queries, and charges only for rows that carry data.

What it collects

  • TRAC searches: every search on your Pulsar subdomain with its name, search hash, status (READY or EDIT), start date and total contents, sorted by name or daily velocity and filtered by search status and realtime status (STARTED, STOPPED, NOT_PRESENT). Paged through the whole list.
  • TRAC search results (mentions): the posts and reactions a search collected, with content, source (X, Reddit, Instagram, Pinterest, Facebook pages, podcasts and the other sources your search covers), author name, publication time, impressions, engagement, likes, comments, shares, views, syndicated copies, potential reach and Pulsar visibility score. Filter by date range, content type (posts or reactions), source, country and hashtag.
  • CORE brands and profiles: the brands you track (owned or competitor) with their brand set and the social profiles inside each, with source and whether the profile is connected.
  • CORE profile metrics: posts, comments, engagements and impressions for a brand's profiles over a date range, summed, averaged or counted, together or one row per profile.
  • CORE profile posts: each post of a profile with its content, comments, likes, engagement and impressions.
  • Read only, pay per result, bring your own token. Nothing in your Pulsar account is created, edited, started or stopped.

Input

FieldWhat it does
What to readTRAC searches (default, needs nothing else), TRAC search results, CORE brands and profiles, CORE profile metrics or CORE profile posts.
Sort searches by, Search status, Realtime statusTRAC searches mode.
Search IDsTRAC search results. The search hashes the TRAC searches mode lists.
Content type, Sources, Countries, HashtagTRAC search results filters.
Brand ID, Profile IDsCORE profile metrics and profile posts. The CORE brands mode lists them.
Aggregation, One row per profileCORE profile metrics: SUM, AVG or COUNT.
Start date, End dateData modes, as YYYY-MM-DD in UTC. Defaults to the last 7 days.
Days per requestData modes read the range in windows of this many days (default 1) because each request returns one page.
Maximum resultsRow cap for the run.
Requests per minutePacing, up to the provider's 5 a second.
API bearer tokenYour own Pulsar API token, as a secret input.

Run TRAC searches first to get search hashes, then TRAC search results with those hashes. For CORE, run CORE brands and profiles first, then profile metrics or profile posts with a brand ID and its profile IDs.

Limits on the Pulsar side

  • API access is part of a Pulsar licence: your account manager enables it per user after Pulsar's product team approves the use case, and the token is issued per user and per subdomain.
  • The API allows 5 requests a second. There is no total call limit.
  • A page holds at most 100 results (CORE post pages hold 50). The provider documents a next token for paging but not how to send it back, so this actor reads one page per search (or profile) per date window. If a window has more results than a page, the log says so: lower Days per request to read more.
  • Some X (Twitter) fields are masked by Pulsar's data contracts: user bio, follower count and coordinates are not available through the API.
  • Only data already indexed in your TRAC searches or CORE brand sets is available.

FAQ

What is Pulsar?

Pulsar (Pulsar Group plc) is an audience intelligence and social listening platform. TRAC is its social listening and audience research product; CORE is its owned and competitor social channel analytics product.

What does this actor read from the Pulsar API?

It sends documented GraphQL read queries to the Pulsar GraphQL API v2.0: the searches query on the TRAC metadata endpoint, the results query on the TRAC data endpoint, and the brands, posts, comments, engagements, impressions and results queries on the CORE data endpoint. It never sends a mutation.

Do I need my own Pulsar API token?

Yes. This is bring your own key. Ask your Pulsar account manager to enable API access for your user, then generate the token on the API token page of your Pulsar account. The same bearer token works on the TRAC and CORE endpoints.

Where do I find a search hash, brand ID or profile ID?

Run the TRAC searches mode for search hashes (also shown on the Search Status page in Pulsar), and the CORE brands and profiles mode for brand and profile IDs.

Can it export more than one page of mentions?

Yes, by date window. Each request returns one page, so a wide date range is split into windows (one day by default, newest first) and each window is read separately. Use a smaller window or a narrower filter for very busy searches.

Does each mention have an ID or URL?

The documented result fields carry no ID or URL, so each row gets a stable hash of its search, date, source, author and content for de-duplication.

What happens if my token is missing or rejected?

The run ends cleanly, with no rows and no charge, and the status message says what to fix. A missing token is caught before any network call.

Example output

{
"mode": "results",
"modeLabel": "TRAC search results",
"recordType": "mention",
"found": true,
"searchId": "88fffa8cf1aa11abc468f7522c71b932",
"brandId": null,
"profileId": null,
"id": "5c1f0e7a2b9d4e8f6a31",
"name": null,
"date": "2024-02-23T15:05:27Z",
"source": "X",
"author": "Example Account",
"content": "Climate change has a negative impact on children...",
"impressions": 5495,
"engagement": 0,
"likes": 0,
"comments": null,
"shares": 0,
"views": null,
"potentialReach": 49956,
"visibility": 91.94,
"syndicationContentsCount": null,
"totalHits": 8925,
"dateFrom": "2024-02-23T00:00:00Z",
"dateTo": "2024-02-23T23:59:59Z",
"retrievedAt": "2026-10-03T09:00:00.000Z",
"note": null
}

Keyword map

  • social listening API, social media listening export, Pulsar API, Pulsar TRAC API, Pulsar CORE API
  • audience intelligence data, brand mentions export, social media monitoring data, share of voice data
  • post engagement metrics, impressions and reach export, competitor social profile analytics
  • GraphQL social listening connector, Power BI or warehouse export of Pulsar data

Source

Built from Pulsar's own API documentation at https://doc.pulsarplatform.com/pulsar-documentation (TRAC Metadata Endpoint, Data Endpoint TRAC, Data Endpoint CORE, Overview and FAQs) and the help centre articles at https://intercom.help/pulsar/en/articles/612893-pulsar-api, checked on 2026-10-03.