How Far Does Free Google Analytics Go? Free vs Paid Explained

Google Analytics (GA4) is free to use. What is less obvious, until you have been running it for a while, is exactly how far the free tier goes and at what point the paid version becomes worth considering.

This article lays out the specific limits that apply to the free tier, based on figures published in Google's official documentation, and explains how those constraints show up in day-to-day work. It then sets out the criteria for deciding whether to move to the paid version, Google Analytics 360. The focus is on the boundaries of the free tier rather than on basic usage or initial setup.

The short answer: most sites are fine on the free tier

For a typical website up to a few hundred thousand pageviews per month, you are unlikely to run into a functional wall on the free tier. There is no charge based on the volume of data collected, and both the standard reports and the Explore feature are available at no cost.

The question is not whether a feature exists, but whether you will hit a ceiling. The specific numbers follow.

Limits that apply to the free tier

The main configuration limits for the free tier (standard properties), as published in Google's official help documentation, are listed below. All of them apply per property.

Configuration limits

• Audiences: 100
• Key events (formerly conversions): 30
• User-scoped custom dimensions: 25
• Event-scoped custom dimensions: 50
• Event-scoped custom metrics: 50
• Item-scoped custom dimensions: 10
• Saved comparisons: 50
• Saved segments: 50
• Custom insights: 50
• Calculated metrics: 5
• Google Ads links: 400

In practice, the two you are most likely to hit first are the 30 key events and the 5 calculated metrics. Five is a particularly tight allowance: if you are tracking several business units or product lines within a single property, you will reach it early.

Data collection limits

• Event parameters per event: 25
• User properties: 25
• Event parameter value length: 100 characters (page_title up to 300, page_referrer up to 420, page_location up to 1,000)
• Distinctly named events: no limit for web data streams; 500 per user for app data streams

The absence of any cap on event volume for web measurement matters. There is a common misconception that measurement stops once traffic grows beyond a certain point, but no such restriction applies to web tracking.

Data retention

Data retention on the free tier is capped at 14 months. The only options are 2 months or 14 months, and the default is 2. A great many accounts are still running on the default, which makes changing this to 14 months one of the first things to do after setup.

Note that this retention period governs the data available in Explore reports. Aggregated figures in the standard reports remain accessible beyond the retention window. If you want to run year-over-year comparisons in Explore, a 2-month setting makes that impossible.

Sampling and export limits

• Explore sampling limit: 10 million events per query
• Maximum events per ad hoc query: 10 million per query (anything beyond this is sampled)
• Report and Explore export: 100,000 rows
• Unsampled explorations: not available on the free tier
• API quota: 200,000 tokens per day
• BigQuery daily export: 1 million events
• Data import storage: 10 GB per property

Three situations where the free tier starts to bind

Here is how those numbers translate into practical constraints, across three typical cases.

1. Sampling degrades the accuracy of your numbers

Sampling is what most analysts run into first. Once an Explore query exceeds 10 million events, the results are calculated from a sample rather than the full dataset.

When sampling occurs, the margin of error grows as segments get narrower. It may not interfere with reading overall trends, but for analysis of the small group of users who converted, it can lead you to the wrong conclusion.

Because unsampled explorations are unavailable on the free tier, avoiding sampling means working around it operationally, by shortening date ranges or narrowing conditions.

2. BigQuery export caps out at 1 million events per day

BigQuery integration is available on the free tier. This is a significant advantage: once you can extract the raw data and aggregate it yourself, most of the constraints imposed by the GA4 interface fall away.

The daily export, however, is capped at 1 million events. Beyond that, the export for that day either fails to run or completes only partially. As a rough guide, at around 10 events per session, 1 million events corresponds to roughly 100,000 sessions per day. For large publishers and e-commerce sites, this is where the practical dividing line sits.

Streaming export, by contrast, has no cap on the free tier. If you are at the scale where the daily export limit is a problem, switching to streaming export is worth evaluating, bearing in mind that BigQuery charges apply separately.

3. Custom dimensions and calculated metrics run out

The more sophisticated your measurement design, the faster you consume custom dimension slots. User-scoped dimensions are especially tight at 25, which tends to fall short when you want to segment membership attributes in any detail.

The 5 calculated metrics limit is easy to overlook. If you want to define your own KPIs within the interface, five will not cover multiple business lines at once.

How Google Analytics 360 differs

The paid version is called Google Analytics 360 and is offered as part of the Google Marketing Platform. The main differences from the free tier are as follows.

Limit comparison

• Event parameters: 25 to 100
• User properties: 25 to 100
• Event-scoped custom dimensions and metrics: 50 each to 125 each
• Item-scoped custom dimensions: 10 to 25
• Key events: 30 to 50
• Audiences: 100 to 400
• Saved comparisons and segments: 50 each to 200 each
• Calculated metrics: 5 to 50
• Data retention: up to 14 months to up to 50 months (2, 14, 26, 38, or 50)
• Explore sampling limit: 10 million events to 1 billion events
• API quota: 200,000 tokens per day to 2 million per day
• BigQuery daily export: 1 million events to billions of events
• Data import storage: 10 GB to 1 TB

Features exclusive to 360

Beyond raised limits, 360 includes capabilities that do not exist on the free tier at all.

• Unsampled explorations: analysis without sampling, within an allowance of 20,000 daily tokens
• Subproperties: child properties created by filtering data from a parent property, useful for separating access by department
• Roll-up properties: multiple properties consolidated into one for cross-property analysis
• SLA: contractual guarantees covering data collection, report freshness, and processing

For organisations running multiple brands or operating across several countries, subproperties and roll-up properties are often a stronger reason to adopt 360 than the higher limits.

Pricing and how to get it

360 is an annual-contract enterprise product, purchased either directly from Google or through a Google Sales Partner. Pricing is not published; it is quoted individually based on data volume and contract terms, so an accurate figure requires a conversation with a sales partner. It is generally understood to run into the millions of yen per year, which is well above the level at which a small or mid-sized site would adopt it purely to relieve functional constraints.

Once contracted, the upgrade is applied per property from the Google Marketing Platform interface. Higher limits take effect immediately after the upgrade, though in some cases it can take up to four hours to apply.

When moving to 360 is worth considering

If any of the following apply, 360 merits evaluation. If none of them do, staying on the free tier is the better call.

1. Sampling occurs every time you use Explore, and the loss of accuracy is interfering with your work
2. The BigQuery daily export has reached, or is approaching, the 1 million event limit
3. You routinely need comparative analysis in Explore across periods longer than 14 months
4. You run multiple brands or business lines and need permission separation via subproperties or cross-property analysis via roll-up properties
5. Custom dimension or calculated metric slots are exhausted, forcing compromises in your measurement design
6. Data gaps or delays feed directly into business decisions, making an SLA necessary

Free alternatives to try first

Even if the above applies, there are options worth checking before migrating to 360.

• Verify that data retention is not still set to 2 months, and change it to 14
• Export to BigQuery and aggregate with SQL rather than in the GA4 interface, which resolves the sampling problem at its root
• Combine with Looker Studio to work around reporting constraints
• Archive unused audiences and custom dimensions to free up slots
• If you have packed several business lines into one property, split it up to distribute the limits

BigQuery integration in particular is available on the free tier and effectively resolves both sampling and retention, which makes it the option to evaluate before considering 360.

Summary

The free tier of Google Analytics carries no charge for data volume and includes BigQuery integration, which makes it sufficient for the majority of sites. The constraints that actually cause problems are confined to a few areas: sampling, the BigQuery daily export cap, the 14-month retention limit, and the allowances for custom dimensions and calculated metrics.

Analytics 360 raises those limits substantially and adds capabilities the free tier lacks, such as unsampled analysis and subproperties. Given that it represents an investment of millions of yen per year, the sensible approach is to exhaust the free workarounds first, such as BigQuery integration and splitting properties, and consider 360 only when a remaining constraint is affecting business decisions.

The limits cited here reflect Google's official documentation as of July 2026. Specifications change, so check the official documentation for current figures before making a decision.