Skip to content

Connector-based Data Mart

Try it free

Use a connector-based Data Mart when you want to collect data directly from a platform API like Facebook Ads, TikTok Ads, LinkedIn Ads or other available data sources (or your own custom connector) into any of the supported Data Storages.

You’ll need a data storage available for the data mart setup. Here is how to add a data storage

Click + New Data Mart.

Give it a friendly Title, e.g., Facebook Ads | Creatives.

Select the Data Storage from the list.

Click Create Data Mart.

Create Data Mart-1

In the Input Source panel, select Definition Type – Connector. Click Setup Connector.

Create Data Mart-2

Select the platform you want to connect to (e.g., Facebook Ads, TikTok Ads) from the list of available connectors and click Next. Connectors built in your project are listed under Custom Connectors; to build one, see Connector Builder.

Create Data Mart-3

Each connector has a set of required parameters. The most common ones include:

ParameterDescription
account_idID of the ad account or business you’re pulling data from
start_dateStart date for the data import
end_dateEnd date for the data import
reimport_lookback_windowHow many days of data to re-import during each run
max_fetching_daysMax days per API call to avoid rate limits
cleanup_to_keep_windowHow many days of data to keep before automatic cleanup

📌 Some connectors may have additional parameters — see their setup guides.

Some parameters use a fixed list of values. Choose these before moving to field selection, especially when the parameter changes the reporting grain. For example, TikTok Ads Data Level changes which fields are required for performance endpoints.

Create Data Mart 4

Each connector requires authentication. Here is how you can obtain the keys for each of the ad platforms and save them in a Credential:

A configuration selects a Credential instead of holding a key. Select one that fits the connector, or click + Create new Credential to add one beside the configuration; it is selected when you save it. With exactly one fitting Credential, a new configuration selects it for you. Create and manage Credentials on the Credentials page; see Source providers.

To set up a Data Mart like one you already have, click Reuse from another Data Mart and pick a Data Mart with the same source. The configuration gets the same Credential and settings, such as account IDs. If that Data Mart’s Credential is not available to you, only the settings are copied, and you select a Credential yourself. No key is copied: Data Marts that share a Credential all get its new values when you replace them once.

Once the connector and parameters are set, click Next to select the node you want to import in this data mart.

Each connector includes one or more data nodes. For example, Facebook Ads has nodes like:

  • ad_accounts
  • ad_accounts/insights
  • ad_accounts/campaigns

👉 Create a separate Data Mart for each node you want to pull data from.

Next, click Select all or select just the fields you want to store in your data warehouse. Required unique-key fields stay selected because the connector needs them to merge data correctly. Some connectors adjust these required fields based on the parameters selected in the previous step.

Create Data Mart 5

Enter dataset name for your data storage (e.g., dataset name for BigQuery) where the connector can load the data. The dataset will be created automatically if it doesn’t exist.

The table name will be created automatically based on the selected node name, for example:
ad_account_adcreatives

Don’t forget to:

  • Click Save & Run at the end of the setup wizard. It saves the settings, publishes the Data Mart and starts the first import.
  • To only save, click Save in the menu next to the button.
  • To start the first import later, click Publish & Run Data Mart in the page header.

To load data again later, use the Manual Run button. It sits in the header of the Input Source card on the Data Setup tab. You can start a run while the card is collapsed.

The Data Setup tab of a published connector Data Mart, with the Manual Run button in the Input Source card header

After the first run, you’ll see that the Output schema was generated automatically.

Each field includes:

  • Name
  • Data type
  • Join key metadata
  • Description

Next, you can add business-friendly names to improve usability later in BI tools as aliases.

Create Data Mart 7

💡 Tip: Add a description to the Data Mart itself so you always know what that is all about.

Create Data Mart 8

Also you can check the Run History tab for logs and results

Choose Backfill (custom period) in the Manual Run sheet to reload a date range from the source.

Pick any past period that starts on or after 2000-01-01. The form shows how many days it covers. A period of up to 31 days is one run. OWOX loads a longer period in consecutive parts, one calendar month per run. The first and last parts cover only the days you picked: 1 January – 15 April loads as 1–31 January, 1–28 February, 1–31 March and 1–15 April. Each part is its own entry in Run History, one after another. Every entry names its backfill period and part, for example “part 3 of 18”.

A part counts as one run for your consumption once any of its accounts loads data. A part that loads nothing, or is cancelled or interrupted, does not count. The form states the number of parts and runs before you start. A period over 31 days asks you to confirm how many runs it costs before it starts.

A failed part never stops the rest of the backfill, so OWOX attempts the whole period. The failed entry shows the last day it loaded, so you can backfill the gap later. If a restart delays a part, OWOX queues it again within 15 minutes.

If the project becomes restricted, the backfill stops. The last entry and the history strip name the parts that did not load. OWOX does not resume the backfill when the restriction lifts, so start a new backfill for those days.

You can cancel a backfill at any time by cancelling its active run. No further parts start, and the parts that already finished stay loaded.

A long backfill can take hours. It uses source API quota and keeps one of the project’s concurrent run slots. OWOX does not start other runs of the same Data Mart, including scheduled runs, until the backfill finishes.

A node that has no date filter, such as a list of ads or campaigns, ignores the period and loads the whole catalog. Its backfill is always one run, whatever period you pick; the form says so and asks for no confirmation.

A connector with no date fields has no period to pick. Examples are Google Sheets and a custom connector with no Time-series node and no Incremental strategy. Its backfill loads all data in one run, the same as an incremental load. If the API of a custom connector filters by date, set the node’s Incremental strategy and date parameters in the Connector Builder, and publish the connector. Then a backfill loads the period you pick, and incremental runs of that node also load by date. A Data Mart pinned to an older version keeps it until you change the version on its Input Source card.

If a deploy or restart interrupts a backfill, the automatic retry resumes it. That retry starts from the day after the last one it fully loaded. It does not reload the days it already imported. Starting a backfill yourself always loads the whole period you choose.

Section titled “Step 6: Set Triggers (Optional but Recommended)”

You can automate the connector by setting a trigger to run it on a schedule.

  1. Go to the Triggers tab and click + Add Trigger
  2. Choose Trigger Type: Connector Run
  3. Set schedule:
    • Daily → choose time & timezone
    • Weekly → select days of the week, time & timezone
    • Monthly → select dates, time & timezone
    • Interval → e.g., every 15 minutes
  • Create reports using this Data Mart
  • Add a destination (Google Sheets, Data Studio, etc.)
  • Explore the data in the Output Schema tab
  • Share it with business users so they can build reports