At this step, you need to define the data sources and their accounts from which the data for attribution modeling should be retrieved.
All required data sources should be connected to Alvie prior to this step. You can learn more about how to connect a data source to Alvie here.
At this step, you need to define two different types of accounts.
1. Define a Google Analytics account
Google Analytics is used as a data source for the attribution model, and it is a required field:
2. Define all other accounts
At least one additional account needs to be defined at this step. These additional accounts serve as a source of data for impressions and costs:
3. Adding custom spend sources
Spend data stored in a BigQuery table or a Google Sheet can be used as additional input for the attribution model. If the spend sources are not supported by the accounts above it’s possible to add additional custom spend sources from a Google sheet or a Bigquery table. These custom spend sources can be added under the Accounts section by selecting the Add Custom Spend Sources button.
Google BigQuery
If your spend data is stored in a BigQuery table, you must first add the corresponding Google Cloud Platform (GCP) project as a data source in your workspace. Once the GCP project is connected, you can select the appropriate Project ID, Dataset ID, and Table Name from the available dropdown menus.
Google Sheets
For spend data stored in a Google Sheet, the sheet must first be added as a data source. Once connected, you can select the sheet from the dropdown menu. Note: selecting specific worksheets is not supported. Please ensure that your spend data is placed in the first worksheet of the Google Sheet.
Data Schema Requirements
Spend data must conform to the following schema, which is validated upon saving the configuration. Any issues will be flagged, allowing you to make corrections based on the validation feedback.
Schema for adding a custom spend source:
date | channel | spend | currency_code |
Bigquery type: DATE
Google sheets type: STRING
The date of the spend for the channel should be in the format "YYYY-MM-DD". | Bigquery type: STRING
Google sheets type: STRING
The channel of the spend. Must match the name (case sensitive) of a Channel in the Custom Grouping | Bigquery type: FLOAT
Google sheets type: FLOAT
The spend value. Strictly non-negative. | Bigquery type: STRING
Google sheets type: STRING
The currency of the spend. Must be ISO 4217. |
Example of a valid Google Sheets table:
Step by step guide to add a Custom spend source:
Best Practices for Custom Spend Sources
Recommended Data Range: To ensure accurate results, include a minimum of 30 days worth of spend data when setting up a custom spend source. This allows the attribution model to analyze sufficient historical trends and deliver better insights.
Limitations
Custom spends sources are only a channel level. This means it’s not possible to optimise budget allocation in Budget Optimiser across campaigns coming from custom sources.
4. Advanced settings: filters for accounts
You can optionally apply filters to the data from the specified accounts. Applying specific filters will allow you to include only a subset of the data from your selected accounts in the model:
You can apply as many filters as required. You won't be able to apply filters on the custom spend sources.
5. Advanced settings: conversion event
At this step, you can also define the Conversion event. This refers to the name of the Google Analytics 4 event that you would like to use as the conversion. You can use 'purchase' to track regular e-commerce revenue:
