The API Report CardAPI Index
THE API REPORT CARD / Software & Data Tools / Google Analytics
Google Analytics

Google Analytics API

Web & Product Analytics · analytics.google.com

GA4 exposes self-serve surfaces: the Data API for reporting, Admin API for configuration, Measurement Protocol for ingestion, and BigQuery export for raw events. All are free on standard properties under OAuth or service accounts. Token-bucket quotas are tight and only GA360 raises them.

Last verified: July 2026Software & Data Tools
API GRADE
A
VERIFIED JUL 2026

SCORECARD

ExistenceGOODFour documented surfaces: Data API, Admin API, Measurement Protocol, and User Deletion, plus BigQuery export.
AccessGOODFully self-serve and free on standard properties; quota ceilings only lift by upgrading to GA360.
CoverageGOODReporting, property configuration CRUD, server-side event ingestion, and raw-event egress via BigQuery.
AuthGOODOAuth 2.0 or Google Cloud service accounts with IAM; Measurement Protocol uses property-scoped secrets.
Docs & DXGOODClient libraries in eight languages and thorough reference docs at developers.google.com/analytics.
StabilityGOOD
MORE FROM THE REPORT CARD
MixpanelAAmplitudeAAbstractAAdobe Creative CloudB+
Supergood: Google Analytics shipped a real API. Most vendors don't; we ship near-native APIs for the rest.

Frequently asked questions

Google Analytics scores A on the API Report Card. GA4 exposes self-serve surfaces: the Data API for reporting, Admin API for configuration, Measurement Protocol for ingestion, and BigQuery export for raw events. All are free on standard properties under OAuth or service accounts. Token-bucket quotas are tight and only GA360 raises them.

Tried to integrate with Google Analytics?
SOURCES
Data API quota is a per-project-per-property-per-hour token bucket (1,250 tokens), and at ~10 tokens per Core request that's only ~125 Core requests/hour for standard properties, 10x more on GA360 but still tight for any real-time dashboard developers.google.com ↗
Token cost is variable and largely opaque: longer date ranges, high-cardinality dimensions (pagePath, custom dimensions), and properties with high event volume burn tokens much faster than the docs imply, leading to surprise 429s easyinsights.ai ↗
2023 quota policy change broke Looker Studio dashboards across the GA4 customer base, many editors using the native GA4 connector could not visualize their data or connect to their data sets and Google explicitly did not consider it a critical issue piwik.pro ↗
Per-property-per-hour quota is roughly 4x the per-project-per-property quota, so a single property only needs to be touched by ~4 different projects (e.g. Looker Studio + Supermetrics + a custom script + Fivetran) before the property-level ceiling is reached and every consumer breaks at once developers.google.com ↗
Looker Studio GA4 connector users get 429 quota errors and report fallback paths require building a BigQuery layer or moving to GA360, both meaningful effort/cost marceldigital.com ↗
Quota limits cannot be raised on standard (free) properties at all, the only self-serve relief is to upgrade the property to GA360 at ~$50k/year, which Google does not present as the answer until customers escalate owox.com ↗
API rate-limit responses don't include a published Retry-After contract; clients have to implement custom backoff and quota-tracking via the includeRateLimitInfo response field that Google has only recently exposed developers.google.com ↗
Concurrent request quota is also bucketed (10 concurrent for standard, 50 for 360), bulk backfill jobs trip this even when total request count is well under the hourly budget developers.google.com ↗
BigQuery Export is a workaround for Data API quotas but introduces a >24h freshness gap (full daily export + intraday streaming with caveats), and the intraday table schema and timing have undergone changes that broke downstream pipelines support.google.com ↗
Standard BigQuery Export is capped at 1M events/day, above that, properties silently stop receiving export rows for the day and customers don't always notice until reports come up short support.google.com ↗
Measurement Protocol events arriving late are only ingested into daily tables for up to 2 calendar days, so server-side ingestion pipelines with retry queues silently drop data on long backlogs ga4bigquery.com ↗
GA4 aggressively samples data and applies hidden 'thresholds' that suppress reporting rows when user counts are low, meaning customers literally cannot see their own data in their own dashboards travis.media ↗
GDPR/CCPA cookie consent requirements mean users who decline cookies are invisible to GA4, sites routinely lose 30-50% of traffic data in the UI vs. server-side ground truth travis.media ↗
15-25% of general visitors block GA outright via ad blockers, and on tech-savvy properties (Hacker News, Reddit, developer audiences) the block rate climbs to ~58% travis.media ↗
GA4 UI was a forced migration from Universal Analytics with a steep relearn curve; reporting paradigms (sessions vs. events, conversions vs. key events) and metric definitions changed in incompatible ways tatvic.com ↗
Standard property data retention is capped at 14 months (user/event data), forcing customers onto BigQuery Export or GA360 for longer-horizon historical analysis analyticscanvas.com ↗
GA4 360 (the only way to materially raise API quotas and unsampled limits) starts at ~$50,000/year and scales by event volume, historically a step down from UA 360 at ~$150k flat but still a hard cliff for mid-market posthog.com ↗
Bot traffic is regularly recorded as real visits despite published filtering, inflating engagement metrics travis.media ↗
Customers don't own the underlying data unless they enable BigQuery Export, Google uses site data across its advertising ecosystem, which is a regulatory and trust concern in the EU travis.media ↗
Only ~12 of the 40+ available event types are used in the average GA4 implementation, suggesting the event model is too complex for the typical small-business marketer it ships to digitalapplied.com ↗
Privacy-first competitors (Plausible, Fathom, Matomo, Simple Analytics, Umami) are growing rapidly specifically because of GA4's cookie/consent gaps, sampling, and Google data-ownership concerns contentsquare.com ↗
© Supergood Systems, Inc.