GA Attribution Scrape

0.2.1 verified Sat May 09 auth: no python

Scrapes attribution data from Google Analytics Model Comparison Tool via JS network requests and sends to BigQuery. Currently v0.2.1, pre-1.0 with sporadic releases. Works only on GA Goals (not ecommerce) and loops through each goal individually.

pip install ga-attribution-scrape

Common errors

error ModuleNotFoundError: No module named 'ga_attribution_scrape' ↓

cause The package is installed under hyphenated name but import uses underscores.

fix

Run pip install ga-attribution-scrape then import using underscores: from ga_attribution_scrape import ScrapeAttribution.

error TypeError: __init__() got an unexpected keyword argument 'sample_size' ↓

cause Older versions (pre-0.2.0) did not accept sample_size parameter.

fix

Update to latest version: pip install --upgrade ga-attribution-scrape

error google.auth.exceptions.DefaultCredentialsError: Could not automatically determine credentials. ↓

cause Credentials path not provided or invalid.

fix

Pass valid credentials_path to ScrapeAttribution or set GOOGLE_APPLICATION_CREDENTIALS environment variable.

Warnings

gotcha Works only on GA Goals, not ecommerce conversions. Does not aggregate goals; loops through each goal individually. ↓

fix Ensure you're using GA Goal IDs, not ecommerce conversion IDs.

gotcha Requires a valid Google Cloud service account key with BigQuery write permissions. The library will fail silently if permissions are missing. ↓

fix Check that the service account has roles/bigquery.dataEditor on the dataset.

breaking The library relies on headless browser automation (likely Selenium) and may break with browser updates. ↓

fix Keep browser drivers updated. Consider pinning browser version in Docker.

Imports

ScrapeAttribution

wrong

from ga_attribution_scrape.scraper import ScrapeAttribution

correct

from ga_attribution_scrape import ScrapeAttribution

Top-level import is correct; submodule path not exposed.

Quickstart

Basic usage: initialize with property ID, goal ID, and credentials, then scrape.

from ga_attribution_scrape import ScrapeAttribution

scraper = ScrapeAttribution(
    property_id='123456789',
    goal_id='90',
    lookback_days=30,
    sample_size=10000,
    credentials_path='path/to/service-account-key.json'
)
results = scraper.scrape()
print(results.head())