Skip to content

Latest commit

 

History

History
383 lines (288 loc) · 45.7 KB

google-ads.md

File metadata and controls

383 lines (288 loc) · 45.7 KB

Google Ads

This page contains the setup guide and reference information for the Google Ads source connector.

Prerequisites

  • (For Airbyte Open Source):
    • A Developer Token
    • OAuth credentials to authenticate your Google account

Setup guide

Step 1: (For Airbyte Open Source) Apply for a developer token

To set up the Google Ads source connector with Airbyte Open Source, you will need to obtain a developer token. This token allows you to access your data from the Google Ads API. Please note that Google is selective about which software and use cases are issued this token. The Airbyte team has worked with the Google Ads team to allowlist Airbyte and ensure you can get a developer token (see issue 1981 for more information on this topic).

  1. To proceed with obtaining a developer token, you will first need to create a Google Ads Manager account. Standard Google Ads accounts cannot generate a developer token.

  2. To apply for the developer token, please follow Google's instructions.

  3. When you apply for the token, make sure to include the following:

    • Why you need the token (example: Want to run some internal analytics)
    • That you will be using the Airbyte Open Source project
    • That you have full access to the code base (because we're open source)
    • That you have full access to the server running the code (because you're self-hosting Airbyte)

:::note You will not be able to access your data via the Google Ads API until this token is approved. You cannot use a test developer token; it has to be at least a basic developer token. The approval process typically takes around 24 hours. :::

Step 2: (For Airbyte Open Source) Obtain your OAuth credentials

If you are using Airbyte Open Source, you will need to obtain the following OAuth credentials to authenticate your Google Ads account:

  • Client ID
  • Client Secret
  • Refresh Token

Please refer to Google's documentation for detailed instructions on how to obtain these credentials.

A single access token can grant varying degrees of access to multiple APIs. A variable parameter called scope controls the set of resources and operations that an access token permits. During the access token request, your app sends one or more values in the scope parameter.

The scope for the Google Ads API is: https://www.googleapis.com/auth/adwords

Each Google Ads API developer token is assigned an access level and "permissible use". The access level determines whether you can affect production accounts and the number of operations and requests that you can execute daily. Permissible use determines the specific Google Ads API features that the developer token is allowed to use. Read more about it and apply for higher access here.

Step 3: Set up the Google Ads connector in Airbyte

For Airbyte Cloud:

To set up Google Ads as a source in Airbyte Cloud:

  1. Log in to your Airbyte Cloud account.
  2. In the left navigation bar, click Sources. In the top-right corner, click + New source.
  3. Find and select Google Ads from the list of available sources.
  4. Enter a Source name of your choosing.
  5. Click Sign in with Google to authenticate your Google Ads account. In the pop-up, select the appropriate Google account and click Continue to proceed.
  6. (Optional) Enter a comma-separated list of the Customer ID(s) for your account. These IDs are 10-digit numbers that uniquely identify your account. To find your Customer ID, please follow Google's instructions. Leaving this field blank will replicate data from all connected accounts.
  7. (Optional) Enter customer statuses to filter customers. Leaving this field blank will replicate data from all accounts. Check Google Ads documentation for more info.
  8. (Optional) Enter a Start Date using the provided datepicker, or by programmatically entering the date in YYYY-MM-DD format. The data added on and after this date will be replicated. (Default start date is 2 years ago)
  9. (Optional) You can use the Custom GAQL Queries field to enter a custom query using Google Ads Query Language. Click Add and enter your query, as well as the desired name of the table for this data in the destination. Multiple queries can be provided. For more information on formulating these queries, refer to our guide below.
  10. (Required for Manager accounts) If accessing your account through a Google Ads Manager account, you must enter the Customer ID of the Manager account.
  11. (Optional) Enter a Conversion Window. This is the number of days after an ad interaction during which a conversion is recorded in Google Ads. For more information on this topic, refer to the Google Ads Help Center. This field defaults to 14 days.
  12. (Optional) Enter an End Date in YYYY-MM-DD format. Any data added after this date will not be replicated. Leaving this field blank will replicate all data from the start date onward.
  13. Click Set up source and wait for the tests to complete.

For Airbyte Open Source:

To set up Google Ads as a source in Airbyte Open Source:

  1. Log in to your Airbyte Open Source account.
  2. In the left navigation bar, click Sources. In the top-right corner, click + New source.
  3. Find and select Google Ads from the list of available sources.
  4. Enter a Source name of your choosing.
  5. Enter the Developer Token you obtained from Google.
  6. To authenticate your Google account, enter your Google application's Client ID, Client Secret, Refresh Token, and optionally, the Access Token.
  7. (Optional) Enter a comma-separated list of the Customer ID(s) for your account. These IDs are 10-digit numbers that uniquely identify your account. To find your Customer ID, please follow Google's instructions. Leaving this field blank will replicate data from all connected accounts.
  8. (Optional) Enter customer statuses to filter customers. Leaving this field blank will replicate data from all accounts. Check Google Ads documentation for more info.
  9. (Optional) Enter a Start Date using the provided datepicker, or by programmatically entering the date in YYYY-MM-DD format. The data added on and after this date will be replicated. (Default start date is 2 years ago)
  10. (Optional) You can use the Custom GAQL Queries field to enter a custom query using Google Ads Query Language. Click Add and enter your query, as well as the desired name of the table for this data in the destination. Multiple queries can be provided. For more information on formulating these queries, refer to our guide below.
  11. (Required for Manager accounts) If accessing your account through a Google Ads Manager account, you must enter the Customer ID of the Manager account.
  12. (Optional) Enter a Conversion Window. This is the number of days after an ad interaction during which a conversion is recorded in Google Ads. For more information on this topic, see the section on Conversion Windows below, or refer to the Google Ads Help Center. This field defaults to 14 days.
  13. (Optional) Enter an End Date in YYYY-MM-DD format. Any data added after this date will not be replicated. Leaving this field blank will replicate all data from the start date onward.
  14. Click Set up source and wait for the tests to complete.

Supported Sync Modes

The Google Ads source connector supports the following sync modes:

Incremental Events Streams

List of stream:

These streams support incremental updates, including deletions, leveraging the Change Status stream. However, they only capture updates from the most recent three months.

The initial sync operates as a full refresh. Subsequent syncs begin by reading updates from the Change Status stream, followed by syncing records based on their IDs.

:::warning It's important to note that the Google Ads API resource ChangeStatus has a limit of 10,000 records per request. That's why you cannot sync stream with more than 10,000 updates in a single microsecond. In such cases, it's recommended to use a full refresh sync to ensure all updates are captured. :::

Supported Streams

The Google Ads source connector can sync the following tables. It can also sync custom queries using GAQL.

Main Tables

Highlights the setup and configurations of a Google Ads account. It encompasses features like call reporting and conversion tracking, giving a clear picture of the account's operational settings and features.

Targeting option for a campaign, such as a keyword, placement, or audience.

Represents the bidding strategy at the campaign level.

Represents labels that can be attached to different entities such as campaigns or ads.

Different attributes of ads from ad groups segmented by date.

Represents an ad group within a campaign. Ad groups contain one or more ads which target a shared set of keywords.

Represents the bidding strategy at the ad group level.

Represents criteria in an ad group, such as keywords or placements.

Represents criteria for listing group ads.

Represents user lists that are defined by the advertiser to target specific users.

A particular interest-based vertical to be targeted.

A click view with metrics aggregated at each click level, including both valid and invalid clicks.

Note that ad_group, ad_group_ad, and campaign contain a labels field, which should be joined against their respective *_label streams if you want to view the actual labels. For example, the ad_group stream contains an ad_group.labels field, which you would join against the ad_group_label stream's label.resource_name field.

Report Tables

Provides in-depth metrics related to ads interactions, including viewability, click-through rates, and conversions. Segments data by various factors, offering a granular look into how ads perform across different contexts.

Represents a campaign in Google Ads.

Represents the budget settings of a campaign.

Geographic View includes all metrics aggregated at the country level. It reports metrics at either actual physical location of the user or an area of interest.

User Location View includes all metrics aggregated at the country level. It reports metrics at the actual physical location of the user by targeted or not targeted location.

Metrics for display keywords, which are keywords that are targeted in display campaigns.

Reporting view that shows metrics aggregated by topic, which are broad categories of interests that users have.

Provides Shopping campaign statistics aggregated at several product dimension levels. Product dimension values from Merchant Center such as brand, category, custom attributes, product condition and product type will reflect the state of each dimension as of the date and time when the corresponding event was recorded.

Provides metrics related to the performance of keywords in the campaign.

Metrics and attributes of legacy ads from ad groups.

:::note Due to Google Ads API constraints, the click_view stream retrieves data one day at a time and can only retrieve data newer than 90 days ago. Also, metrics cannot be requested for a Google Ads Manager account. Report streams are only available when pulling data from a non-manager account. :::

:::warning Google Ads doesn't support PERFORMANCE_MAX campaigns on ad_group or ad stream level, only on campaign level. If you have this type of campaign Google will remove them from the results for the ads reports. More info and Google Discussions. :::

For incremental streams, data is synced up to the previous day using your Google Ads account time zone since Google Ads can filter data only by date without time. Also, some reports cannot load data real-time due to Google Ads limitations.

Reasoning Behind Primary Key Selection

Primary keys are chosen to uniquely identify records within streams. In this selection, we considered the scope of ID uniqueness as detailed in the Google Ads API structure documentation. This approach guarantees that each record remains unique across various scopes and contexts. Moreover, in the Google Ads API, segmentation is crucial for dissecting performance data. As pointed out in the Google Ads support documentation, segments offer a granular insight into data based on specific criteria, like device type or click interactions.

Custom Query: Understanding Google Ads Query Language

Additional streams for Google Ads can be dynamically created using custom queries.

The Google Ads Query Language queries the Google Ads API. Review the Google Ads Query Language and the query builder to validate your query. You can then add these as custom queries when configuring the Google Ads source.

Example GAQL Custom Query:

SELECT
    campaign.name,
    metrics.conversions,
    metrics.conversions_by_conversion_date
FROM ad_group

Note that segments.date is automatically added to the WHERE clause if it is included in the SELECT clause. Custom reports including segments.date in the SELECT clause will be synced by day.

Each custom query in the input configuration must work for all the customer account IDs. Otherwise, the customer ID will be skipped for every query that fails the validation test. For example, if your query contains metrics fields in the select clause, it will not be executed against manager accounts.

Follow Google's guidance on Selectability between segments and metrics when editing custom queries or default stream schemas (which will also be turned into GAQL queries by the connector). Fields like segments.keyword.info.text, segments.keyword.info.match_type, segments.keyword.ad_group_criterion in the SELECT clause tell the query to only get the rows of data that have keywords and remove any row that is not associated with a keyword. This is often unobvious and undesired behavior and can lead to missing data records. If you need this field in the stream, add a new stream instead of editing the existing ones.

:::info For an existing Google Ads source, when you are updating or removing Custom GAQL Queries, you should also subsequently refresh your source schema to pull in any changes. :::

Difference between manager and client accounts

A manager account isn't an "upgrade" of your Google Ads account. Instead, it's an entirely new Google Ads account you create. Think of a manager account as an umbrella Google Ads account with several individual Google Ads accounts linked to it. You can link new and existing Google Ads accounts, as well as other manager accounts.

You can then monitor ad performance, update campaigns, and manage other account tasks for those client accounts. Your manager account can also be given ownership of a client account. This allows you to manage user access for the client account.

Link for more details on how it works and how you can create it.

Manager Accounts (MCC) primarily focus on account management and oversight. They can access and manage multiple client accounts, view shared resources, and handle invitations to link with client accounts.

Client Accounts are more operationally focused. They deal with campaign management, bidding, keywords, targeting, extensions, metrics, reporting, billing, and other ad-specific functionalities.

While both types of accounts can access a wide range of resources in the API, the difference lies in their scope and purpose. Manager accounts have a broader oversight, while client accounts delve into the specifics of advertising operations.

For detailed information, refer to the official documentation.

Note on Conversion Windows

In digital advertising, a 'conversion' typically refers to a user undertaking a desired action after viewing or interacting with an ad. This could be anything from clicking through to the advertiser's website, signing up for a newsletter, making a purchase, and so on. The conversion window is the period of time after a user sees or clicks on an ad during which their actions can still be credited to that ad.

For example, imagine an online shoe store runs an ad and sets a conversion window of 30 days. If you click on that ad today, any purchases you make on the shoe store's site within the next 30 days will be considered conversions resulting from that ad. The length of the conversion window can vary depending on the goals of the advertiser and the nature of the product or service. Some businesses might set a shorter conversion window if they're promoting a limited-time offer, while others might set a longer window if they're advertising a product that consumers typically take a while to think about before buying.

In essence, the conversion window is a tool for measuring the effectiveness of an advertising campaign. By tracking the actions users take after viewing or interacting with an ad, businesses can gain insight into how well their ads are working and adjust their strategies accordingly.

In the case of configuring the Google Ads source connector, each time a sync is run the connector will retrieve all conversions that were active within the specified conversion window. For example, if you set a conversion window of 30 days, each time a sync is run, the connector will pull all conversions that were active within the past 30 days. Due to this mechanism, it may seem like the same campaigns, ad groups, or ads have different conversion numbers. However, in reality, each data record accurately reflects the number of conversions for that particular resource at the time of extracting the data from the Google Ads API.

Performance considerations

This source is constrained by the Google Ads API limits

Due to a limitation in the Google Ads API which does not allow getting performance data at a granularity level smaller than a day, the Google Ads connector usually pulls data up until the previous day. For example, if the sync runs on Wednesday at 5 PM, then data up until Tuesday midnight is pulled. Data for Wednesday is exported only if a sync runs after Wednesday (for example, 12:01 AM on Thursday) and so on. This avoids syncing partial performance data, only to have to resync it again once the full day's data has been recorded by Google. For example, without this functionality, a sync which runs on Wednesday at 5 PM would get ads performance data for Wednesday between 12:01 AM - 5 PM on Wednesday, then it would need to run again at the end of the day to get all of Wednesday's data.

Changelog

Version Date Pull Request Subject
3.3.3 2024-02-14 35280 Temporary patch that disables some fields to avoid 500 error when syncing user_interest steam
3.3.2 2024-02-12 35158 Manage dependencies with Poetry.
3.3.1 2024-01-16 34007 prepare for airbyte-lib
3.3.0 2024-01-12 34212 Remove metric from query in Ad Group stream for non-manager account
3.2.1 2024-01-12 34200 Disable raising error for not enabled accounts
3.2.0 2024-01-09 33707 Add possibility to sync all connected accounts
3.1.0 2024-01-09 33603 Fix two issues in the custom queries: automatic addition of segments.date in the query; incorrect field type for DATE fields.
3.0.2 2024-01-08 33494 Add handling for 401 error while parsing response. Add metrics.cost_micros field to Ad Group stream.
3.0.1 2023-12-26 33769 Run a read function in a separate thread to enforce a time limit for its execution
3.0.0 2023-12-07 33120 Upgrade API version to v15
2.0.4 2023-11-10 32414 Add backoff strategy for read_records method
2.0.3 2023-11-02 32102 Fix incremental events streams
2.0.2 2023-10-31 32001 Added handling (retry) for InternalServerError while reading the streams
2.0.1 2023-10-27 31908 Base image migration: remove Dockerfile and use the python-connector-base image
2.0.0 2023-10-04 31048 Fix schem default streams, change names of streams.
1.0.0 2023-09-28 30705 Fix schemas for custom queries
0.11.1 2023-09-26 30758 Exception should not be raises if a stream is not found
0.11.0 2023-09-23 30704 Update error handling
0.10.0 2023-09-19 30091 Fix schemas for correct primary and foreign keys
0.9.0 2023-09-14 28970 Add incremental deletes for Campaign and Ad Group Criterion streams
0.8.1 2023-09-13 30376 Revert pagination changes from 0.8.0
0.8.0 2023-09-01 30071 Delete start_date from required parameters and fix pagination
0.7.4 2023-07-28 28832 Update field descriptions
0.7.3 2023-07-24 28510 Set dates with client's timezone
0.7.2 2023-07-20 28535 UI improvement: Make the query field in custom reports a multi-line string field
0.7.1 2023-07-17 28365 0.3.1 and 0.3.2 follow up: make today the end date, not yesterday
0.7.0 2023-07-12 28246 Add new streams: labels, criterions, biddig strategies
0.6.1 2023-07-12 28230 Reduce amount of logs produced by the connector while working with big amount of data
0.6.0 2023-07-10 28078 Add new stream Campaign Budget
0.5.0 2023-07-07 28042 Add metrics & segment to Campaigns stream
0.4.3 2023-07-05 27959 Add audience and user_interest streams
0.3.3 2023-07-03 27913 Improve Google Ads exception handling (wrong customer ID)
0.3.2 2023-06-29 27835 Fix bug introduced in 0.3.1: update query template
0.3.1 2023-06-26 27711 Refactor date slicing; make start date inclusive
0.3.0 2023-06-26 27738 License Update: Elv2
0.2.24 2023-06-06 27608 Improve Google Ads exception handling
0.2.23 2023-06-06 26905 Replace deprecated authSpecification in the connector specification with advancedAuth
0.2.22 2023-06-02 26948 Refactor error messages; add pattern_descriptor for fields in spec
0.2.21 2023-05-30 25314 Add full refresh custom table asset_group_listing_group_filter
0.2.20 2023-05-30 25624 Add asset Resource to full refresh custom tables (GAQL Queries)
0.2.19 2023-05-15 26209 Handle Token Refresh errors as config_error
0.2.18 2023-05-15 25947 Improve GAQL parser error message if multiple resources provided
0.2.17 2023-05-11 25987 Categorized Config Errors Accurately
0.2.16 2023-05-10 25965 Fix Airbyte date-time data-types
0.2.14 2023-03-21 24945 For custom google query fixed schema type for "data_type: ENUM" and "is_repeated: true" to array of strings
0.2.13 2023-03-21 24338 Migrate to v13
0.2.12 2023-03-17 22985 Specified date formatting in specification
0.2.11 2023-03-13 23999 Fix incremental sync for Campaigns stream
0.2.10 2023-02-11 22703 Add support for custom full_refresh streams
0.2.9 2023-01-23 21705 Fix multibyte issue; Bump google-ads package to 19.0.0
0.2.8 2023-01-18 21517 Write fewer logs
0.2.7 2023-01-10 20755 Add more logs to debug stuck syncs
0.2.6 2022-12-22 20855 Retry 429 and 5xx errors
0.2.5 2022-11-22 19700 Fix schema for campaigns stream
0.2.4 2022-11-09 19208 Add TypeTransofrmer to Campaings stream to force proper type casting
0.2.3 2022-10-17 18069 Add segments.hour, metrics.ctr, metrics.conversions and metrics.conversions_values fields to campaigns report stream
0.2.2 2022-10-21 17412 Release with CDK >= 0.2.2
0.2.1 2022-09-29 17412 Always use latest CDK version
0.2.0 2022-08-23 15858 Mark the query and table_name fields in custom_queries as required
0.1.44 2022-07-27 15084 Fix data type ad_group_criterion.topic.path in display_topics_performance_report and shifted campaigns to non-managers streams
0.1.43 2022-07-12 14614 Update API version to v11, update google-ads to 17.0.0
0.1.42 2022-06-08 13624 Update google-ads to 15.1.1, pin protobuf==3.20.0 to work on MacOS M1 machines (AMD)
0.1.41 2022-06-08 13618 Add missing dependency
0.1.40 2022-06-02 13423 Fix the missing data issue
0.1.39 2022-05-18 12914 Fix GAQL query validation and log auth errors instead of failing the sync
0.1.38 2022-05-12 12807 Documentation updates
0.1.37 2022-05-06 12651 Improve integration and unit tests
0.1.36 2022-04-19 12158 Fix *_labels streams data type
0.1.35 2022-04-18 9310 Add new fields to reports
0.1.34 2022-03-29 11602 Add budget amount to campaigns stream.
0.1.33 2022-03-29 11513 When end_date is configured in the future, use today's date instead.
0.1.32 2022-03-24 11371 Improve how connection check returns error messages
0.1.31 2022-03-23 11301 Update docs and spec to clarify usage
0.1.30 2022-03-23 11221 Add *_labels streams to fetch the label text rather than their IDs
0.1.29 2022-03-22 10919 Fix user location report schema and add to acceptance tests
0.1.28 2022-02-25 10372 Add network fields to click view stream
0.1.27 2022-02-16 10315 Make ad_group_ads and other streams support incremental sync.
0.1.26 2022-02-11 10150 Add support for multiple customer IDs.
0.1.25 2022-02-04 9812 Handle EXPIRED_PAGE_TOKEN exception and retry with updated state.
0.1.24 2022-02-04 9996 Use Google Ads API version V9.
0.1.23 2022-01-25 8669 Add end date parameter in spec.
0.1.22 2022-01-24 9608 Reduce stream slice date range.
0.1.21 2021-12-28 9149 Update title and description
0.1.20 2021-12-22 9071 Fix: Keyword schema enum
0.1.19 2021-12-14 8431 Add new streams: Geographic and Keyword
0.1.18 2021-12-09 8225 Include time_zone to sync. Remove streams for manager account.
0.1.16 2021-11-22 8178 Clarify setup fields
0.1.15 2021-10-07 6684 Add new stream click_view
0.1.14 2021-10-01 6565 Fix OAuth Spec File
0.1.13 2021-09-27 6458 Update OAuth Spec File
0.1.11 2021-09-22 6373 Fix inconsistent segments.date field type across all streams
0.1.10 2021-09-13 6022 Annotate Oauth2 flow initialization parameters in connector spec
0.1.9 2021-09-07 5302 Add custom query stream support
0.1.8 2021-08-03 5509 Allow additionalProperties in spec.json
0.1.7 2021-08-03 5422 Correct query to not skip dates
0.1.6 2021-08-03 5423 Added new stream UserLocationReport
0.1.5 2021-08-03 5159 Add field login_customer_id to spec
0.1.4 2021-07-28 4962 Support new Report streams
0.1.3 2021-07-23 4788 Support main streams, fix bug with exception DATE_RANGE_TOO_NARROW for incremental streams
0.1.2 2021-07-06 4539 Add AIRBYTE_ENTRYPOINT for Kubernetes support
0.1.1 2021-06-23 4288 Fix Bugfix: Correctly declare required parameters