Skip to main content
POST
Brand Sentiment Analysis API

Brand Sentiment Analysis

One number answers “how does the web feel about us”, and this endpoint gives you the breakdown behind it. Pages mentioning your query are grouped by their main sentiment and, separately, by their strongest emotion: happiness, love, fun, share, sadness and anger. Each group reports how many pages it holds and the sites, countries, languages and site types that contribute most, so you can tell whether criticism comes from news outlets or from a single forum.
Provide a brand, product or topic in query (up to 100 characters). Each request costs 10 credits and returns one row. Very broad terms with millions of mentions take longer to analyze, up to a few minutes.

Standard Result Fields

Task results use the stable SocQ fields below. Source-specific field names and formats are normalized without inferring or enriching values that the source did not provide. To request a smaller field set, use fields for MCP queries, _result_fields for typed tools, or --result-fields in the CLI. Declared fields remain present when their value is null. The following example shows the response format for a completed task. Actual data depends on your query.

Completed Task Response

Authorizations

Authorization
string
header
required

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Body

application/json

Provide a brand, product or topic in query. The result is one row splitting the web pages that mention it by sentiment (positive, negative, neutral) and by emotion, with the leading sites and countries for each.

query
string
required

Brand, product or topic to look for. Wrap a phrase in double quotes to match it exactly.

Minimum string length: 1
callback_url
string<uri>

Response

Task accepted

code
integer
data
object