- All API endpoints require key authentication. Please include the parameter 'api_key' with each API request. The api_key parameter should always be set as a URI query string parameter. Authorization failures result in an HTTP status code of 403
- Due to size limitations on HTTP GET requests, some GET endpoints also support POST requests and are denoted with an asterisk (*)
- Some request parameters are marked experimental and could be subject to change, deprecation, or removal. Accuracy of results cannot be guaranteed when using any experimental features
- API query results should be cached whenever possible. For example, influencer profiles don't vary much from day to day, so caching the output of influencers/show for at least 24h will allow better use of API quotas (fewer calls) and limit the load on the search engine to ensure fast response times
- For any questions regarding the API, please contact email@example.com
- Most common API request failures are a result of missing/blank required parameters or malformed/invalid parameters. These failures will result in an HTTP status code of 400
- Unexpected server failures will result in an HTTP status code of 500
- Please check individual endpoints for any additional error handling
- All API responses are in JSON format. By default, the JSON is compacted. To enable pretty JSON in the response , include the parameter 'pretty' as a URI query string parameter in your API request
- For POST API requests, the Content-Type header needs to be set to 'application/x-www-form-urlencoded; charset=utf-8'
- To enable GZIP compression for an API response, include a value of 'gzip' in the Accept-Encoding request header
- There are few different types of keywords: Standard content keywords, link keywords, and Lucene-style content keywords
- Standard content keywords are searched against the influencer's content and can be either regular search terms (e.g.
traackr) or phrases (e.g.
"influencer marketing"). Phrases must be double quoted - single quotes will be treated as apostrophes or discarded
- Link keywords are searched against embedded links found in the influencer's content and should be prefixed with the label 'link:'. A link keyword can either denote an exact search (e.g.
link:foo.com/post/1 - match content with that embedded URL) or a prefix search (e.g.
link:^foo.com - match content with embedded URLs starting with that domain)
- Lucene-style content keywords are made up of standard content keywords and also search against an influencer's content, but provide support for more advanced Boolean logic
- There is limited boolean query support using standard keywords: keyword terms delimited by a space are ANDed together and comma-separated keyword terms are ORed together. E.g.:
influencer marketing,content marketing is equivalent to
(influencer AND marketing) OR (content AND marketing)
- For more complex boolean queries, Lucene-style expressions using the
+ operators should be used. E.g.:
((influencer AND marketing) OR "influencer marketing strategy" NOT "advertising")
- Any Lucene-style keyword must escape special characters within search terms. The characters to be escaped are
/. The characters from this set that may be used as operators are
" " for phrase queries,
( ) for precedence, and
+ - as boolean MUST and MUST_NOT
- To signify that a keyword is a Lucene expression, it must be wrapped in parentheses - the first character must be a left parenthesis and the last character must be a right parenthesis
- The following characters are illegal for use in content keywords:
- Link keywords MUST NOT be used inside Lucene expressions, as the syntax is not compatible
- Lucene-style keywords should be used only in the keywords parameter, not exclusion_keywords. Any exclusions can be expressed using