OrcLink
← Help Center

Data out

BI tools: Tableau, Power BI, Looker, Domo

Read report rows and daily ad spend with the read API; starter queries for Power BI and a CSV script.

Two ways in:

  • The warehouse export (Data out > Warehouse): OrcLink writes the tables outcomes, touchpoints, credits, spend and calls, one UTC day at a time, to Amazon S3, Google BigQuery, PostgreSQL or Microsoft SQL Server (Snowflake and Redshift read the S3 export with a script OrcLink generates). Your BI tool reads your warehouse. It runs once a day while the scheduled jobs export and jobs run (your OrcLink operator schedules them). Files are CSV (gzip); Parquet is not built. The columns are published at /docs/export-schema.json. See the help pages of each destination.
  • The read API: report rows and daily ad spend, as JSON. The API reference is at /docs/api. The rest of this page is about the read API.

Before you start

  • Make an API key with the read scope only (Data out > API).
  • Send it as a header: Authorization: Bearer orc_live_.... A key in the URL is refused (400 key_in_url), so the tool must be able to send a header.
  • Pages: up to 200 rows per page; pass next_cursor as cursor until it is null.
  • Limits per key: 120 reads and 30 reports per minute; one performance report per key at a time.
  • Credited numbers are OrcLink attribution: a claim. The source system owns each outcome and its value. Money is per currency, never converted.
  • In HIPAA mode, report rows are by channel only and campaign names are left out, unless the owner turned on "Include campaign names in data out".

Useful endpoints:

DataEndpoint
Attribution report rows (by channel, source, campaign, campaign group, ad group or keyword; a rule-based model)GET /api/v1/reports/performance?from=YYYY-MM-DD&to=YYYY-MM-DD&dimension=channel&model=last_non_direct
Daily ad spend (platform, campaign, ad group, keyword, impressions, clicks)GET /api/v1/spend?from=YYYY-MM-DD&to=YYYY-MM-DD

Power BI

Starter queries (Power Query M, plain text): docs/guides/bi/orclink-spend.pq and docs/guides/bi/orclink-performance.pq. In Power BI Desktop: Get data > Blank query > Advanced editor, paste one, set ApiKey (better: a parameter), From and To. When Power BI asks how to connect to the web source, choose Anonymous: the key travels in the header the query sets.

Tableau

Tableau has no built-in way to page a JSON API that needs a header. Options:

  • Connect Tableau to your warehouse (the warehouse export).
  • Run the starter script docs/guides/bi/orclink_to_csv.py (Python 3, no extra packages) on a schedule; it writes orclink_spend.csv and orclink_performance.csv. Point Tableau at the files (or a shared folder).
  • Build a Web Data Connector with Tableau's own SDK (not provided by OrcLink).

Looker (Google Cloud)

Looker reads SQL databases, not web APIs. Use the warehouse export to BigQuery, PostgreSQL or SQL Server (or Snowflake / Redshift over the S3 export) and point Looker at it; or load the CSV files of the script above into your database.

Domo

Use a Domo connector for your warehouse (the warehouse export), one that reads a JSON REST API with a custom header and cursor paging, or upload the CSV files from the script. OrcLink has not checked any Domo connector.