Marin Software

Marin Software is an online advertising platform to measure, manage, and optimize ad spending across the web and mobile devices. Marin provides a Data Passback service which exports reporting data to an SFTP server. Fivetran syncs files from Marin's Data Passback SFTP server to your destination.

Features

Supported deployment models

We support the SaaS Deployment model for the connector.

Source API version

This connector uses JCraft version 0.1.55.

Setup guide

Follow our step-by-step Marin Software setup guide to connect Marin Software with your destination using Fivetran connectors.

Sync overview

• The Manifest file is a summary of all files generated for the given day/account and indicates that the batch of files is ready for processing by the customer's import process. Filenames present in manifest file will be synced. Files that are not in the manifest will be skipped.

Schema information

This schema applies to all Marin Software connections.

Schema notes

Marin maintains an archive of the output files for 7 days from the generation date. If you pause your connection for more than 7 days then you may lose data older than 7 days.

Setting tables

Setting tables represents settings within the application. ACCOUNT_HISTORY, PUBLISHER_HISTORY, CAMPAIGN_HISTORY, GROUP_HISTORY, KEYWORD_HISTORY, CREATIVE_HISTORY, PLACEMENT_HISTORY, PRODUCT_TARGET_HISTORY, CONVERSION_EVENT, DIMENSION and TRACKING_VALUE are the setting tables.

• Publisher accounts denominated in other currencies are converted to the Marin Client Account’s currency. Original currency values are not available in the Data Passback.

• Search objects (CAMPAIGN, GROUP, KEYWORD, etc.) that have been deleted from the publisher prior to linking are not tracked in Marin and are not present in the Data Passback.

• Notes about DIMENSION table,

  • The DIMENSION table has Customer-defined dimensions (meta-data).
  • DIMENSION table records are a means to attach meta-data to any object in the hierarchy. So, for example, if you want to associate a campaign as a "branded" or "non-branded" campaign, you can set up a dimension called “Branded” and tag campaigns with these values.

• Notes about TRACKING_VALUE table,

  • It is possible for one record to concurrently have multiple tracking values.

Fact tables

Fact tables have daily metrics (impressions, clicks, conversions, etc.). KEYWORD_FACT, CREATIVE_FACT, PLACEMENT_FACT, PRODUCT_TARGET_FACT and GROUP_DEVICE_FACT are the fact tables.

• Metrics reflect the date boundaries used by the publisher and/or conversion data sources, which may be different from each other. Publishers will always send data adjusted to the time zone of the publisher account. However, conversion sources are typically fixed in one time zone, which may be different from the publisher account time zone. Data should match the Marin Enterprise User Interface, but is not guaranteed to match Third-party sources.

• Information about keywords without impressions can be derived from the list of all keywords by excluding those keywords for which metrics are present.

• To aggregate metrics, rows must be summed together, substituting zero for null values.

• Marin aggregates traffic and revenue data by KEYWORD and by CREATIVE separately; it does not support a (by KEYWORD) X (by CREATIVE) combination.

• Notes about KEYWORD_FACT table,

  • GROUP, CAMPAIGN, PUBLISHER and ACCOUNT level data can be rolled up from KEYWORD_FACT alone, allowing PLACEMENT_FACT and CREATIVE_FACT records to be disregarded for aggregation purposes. For convenience, unattributed, creative attribution, placement attribution are represented in the table records as below:

    • keyword_id = 0 for content network activity within a GROUP
    • keyword_id = -1 for unattributed conversions within a GROUP
    • keyword_id = -2 for placement conversions within a GROUP
    • keyword_id = -3 for product target conversions within a GROUP
    • keyword_id (when not 0, -1, -2 or -3) is unique number across client accounts

  • Publisher metrics (search_bid, impressions, clicks, publisher cost, avg. position, avg cpc) appear on conversion_event_id = 1 i.e. Total conversions (date of click)

  • The combination of keyword_id = 0 & group_id = 0 represents deleted groups in Marin, where the fact data persists. The data is being passed in the Data Passback, so it can be preserved if relevant for historical data. The group_id = 0 will not exist in the GROUP_HISTORY table and shouldn’t be considered when factoring in referential integrity.

  • To calculate the GROUP and CAMPAIGN level Impression Share metrics displayed in the application, roll-up on the group_id, selecting a keyword_id having a value not equal to 0, -1, -2 or -3 and perform the following calculation:

    • Impr. Share % = (impressions / potential_impressions) * 100

• Notes about GROUP_DEVICE_FACT table,

  • Publisher metrics (impressions, clicks, publisher cost) appear on conversion_event_id = 1 i.e. Total conversions (date of click)
  • The group_id = 0 represents deleted groups in Marin, where the fact data persists. The data is being passed in the Data Passback, so it can be preserved if relevant for historical data. The group_id = 01 will not exist in the GROUP_HISTORY table and shouldn’t be considered when factoring in referential integrity.
  • To calculate the GROUP and CAMPAIGN level Impression Share metrics displayed in the application, roll-up on the group_id, selecting a keyword_id having a value not equal to 0, -1, -2 or -3 and perform the following calculation:
    • Impr. Share % = (impressions / potential_impressions) * 100