7 Google Analytics Alternatives for 2026

Most people looking for a Google Analytics alternative are not reacting to a headline. They hit a specific wall: a report that stopped at 14 months, a number that did not match Shopify, a table where half the revenue landed in a row called other. This guide documents GA4's real free-tier limits from Google's own help pages, then compares seven alternatives on prices read from each vendor's own site in September 2026. One of them, Arktis, is ours, and the section says so.

The short answer: Arktis for stores, Plausible for simple traffic, Matomo if the data must stay yours

If you run an online store, Arktis is the clear winner and the reason is specific. GA4 removed the first click, linear, time decay and position-based attribution models in November 2023, leaving data-driven and two last-click models, so the model comparison a merchant needs is no longer in the product at any price. Arktis captures the fbclid and gclid identifiers on landing, matches Shopify orders and Stripe customers back to the visitor, and reports customer acquisition cost and return on ad spend per campaign across all five models, with the buyer's session recording next to the number. $49, $149 and $349 a month, every tier published, 7-day free trial on Growth.

If you want a clean traffic dashboard without a consent banner, Plausible starts at $9 a month and is hosted in the EU. If a policy says the data stays in your infrastructure, Matomo self-hosted is free forever.

What GA4 actually limits on the free tier

These are from Google's own documentation, not from a competitor's marketing page, and they are the four that send people looking.

Retention. Event-level data retention on a standard property can be set to 2 or 14 months, and 14 is the ceiling, with large properties forced down to 2. The 26, 38 and 50-month options are Analytics 360 only and cover event data; user-level and key-event data cap at 14 months on 360 too. The nuance most articles miss is that this cap applies to explorations and funnel reports, not to standard aggregated reports, so the wall appears the moment you try to analyse something rather than glance at it.

Sampling. Ad-hoc queries are sampled above 10 million events on a standard property, against 100 million by default on 360, and up to 1 billion when you ask Explore for more detailed results. Google's own stated remedy for free-tier users is to export to BigQuery for unsampled results or to narrow the date range.

Cardinality. When a table exceeds its row limit, less common values are condensed into a row called other. Google treats any dimension with more than 500 daily values as high cardinality and names item ID, page path and page location as the usual offenders, which is exactly the shape of an e-commerce catalogue. Google does not publish the per-table row limits that trigger the (other) row.

BigQuery export. The free daily export is capped at one million events a day, and Google warns that a property which consistently exceeds it will have the export paused, with previous days not reprocessed. Streaming export has no volume limit but bills at $0.05 per gigabyte, roughly 600,000 events.

Analytics 360 lifts all four. Google has never published its price; Cardinal Path, a Google Marketing Platform sales partner, put the suggested retail starting price at $50,000 a year including 25 million monthly events, in August 2022. Figures circulating in the $150,000 range generally refer to the older Universal Analytics 360 contract.

The attribution problem, which is the real reason stores leave

In November 2023 Google removed the first click, linear, time decay and position-based attribution models from GA4. What remains is data-driven attribution, paid and organic last click, and Google paid channels last click. Model data does not predate 1 November 2021 for data-driven or 14 June 2021 for the rules-based models, so there is no Universal Analytics history behind it, and all remaining models exclude direct visits from credit unless the entire path was direct.

For a merchant this is not an inconvenience, it is the end of the exercise. The useful question is how much your answer changes when you switch models, because a channel that looks profitable on last click and terrible on first touch is telling you something. GA4 can no longer show you that comparison.

The data loss compounds it. Didomi's European consent benchmark, collected through 2025, puts the Western Europe opt-in rate at 55.7 percent of banners displayed. GWI's Q2 2025 data puts ad blocker use at 29.5 percent of internet users aged 16 to 64, at least sometimes when online. Safari has capped JavaScript-set cookies at seven days since 2019, and cuts that to 24 hours when the visitor arrives on a link carrying a click identifier from a domain it classifies as a tracker, which is exactly what an ad click looks like. Littledata, which sells the fix so weight it accordingly, reports that around 20 of every 100 Shopify orders fail to appear in Google Analytics. GA4's behavioural modelling can fill some of that gap, but only above thresholds most stores never reach: at least 1,000 events a day from users who declined consent for seven days, and at least 1,000 daily consented users on seven of the previous 28 days.

The 7 best Google Analytics alternatives

1. Arktis, the winner for online stores

Arktis is built around the question GA4 stopped answering. It captures ad click identifiers when a visitor lands, holds them across sessions, matches a Stripe customer back to that visitor through a four-pass waterfall and a Shopify order on email, campaign and checkout token, then reports customer acquisition cost and return on ad spend per campaign with first-touch, last-touch, linear, time-decay and position-based models side by side. Session recordings of the attributed buyers sit next to the revenue, gated behind explicit consent, with capture buffered in the browser and discarded unless consent is given, and data hosted in the EU by default.

Pricing is on monthly tracked visitors: $49 for 50,000, $149 for 250,000 with recordings and custom models, $349 for unlimited with API access and raw export, plus a 7-day trial on Growth. It is not a general-purpose web analytics tool, and it does not try to be one: if you want pageview reporting for a content site, Plausible or Matomo will serve you better.

2. Plausible, the simple EU-hosted swap

Plausible is one script, one page, no cookie banner required, and it is open source with a self-hostable community edition. Starter is $9 a month or $90 a year at 10,000 monthly pageviews, Growth $14 or $140 for up to three sites, Business $19 or $190 for ten sites with funnels, e-commerce revenue attribution and five years of retention. There is no free tier, but the 30-day trial needs no card and runs on Business limits. All visitor data is processed on European-owned infrastructure.

It attributes revenue by channel on Business, which is genuinely useful, but it does not ingest ad spend, so it cannot produce a cost per customer.

3. Matomo, for ownership

Self-hosted Matomo Community is free forever with unlimited hits and users, which makes it the only tool here that can be both free and fully yours. The paid plugin bundles add the interesting parts: Team at 2,750 euro a year brings funnels, users flow, heatmaps and session recording; Business at 14,500 euro adds cohorts, multi-channel conversion attribution and advertising conversion export. Matomo Cloud stores data in Europe and starts at about 22 euro a month for 50,000 hits, though Matomo prices by country with no currency selector.

Its attribution runs in the opposite direction to what a merchant wants: it exports conversions to ad platforms rather than importing spend, so there is no native customer acquisition cost.

4. Microsoft Clarity, free behaviour data

Clarity is free with no paid tier and no volume cap, and it gives you heatmaps, session recordings and rage-click detection that GA4 does not have at all. Recordings are kept for 30 days, with favourited recordings and aggregated click data kept nine months. It is not a GA4 replacement for traffic reporting, it is the other half of the picture, and running it alongside something else is the common pattern.

5. PostHog, for product teams who want everything in one stack

PostHog bundles analytics, session replay, feature flags, surveys and experiments, gives away one million events and 5,000 recordings a month, then charges $0.000198 per identified event and $0.00005 per anonymous one. It is open source and self-hostable. The interface assumes technical fluency, and four separate meters mean four ways to be surprised by a bill.

6. Mixpanel, for event analytics at a known rate

If your questions are about what users do rather than where they came from, Mixpanel answers them better than GA4 ever did, and it publishes its rate: one million events a month free, then $0.00028 per event, with 20,000 session replays included on Growth. It charges the same for anonymous and identified events, which matters if most of your traffic never signs in.

7. Amplitude, for enterprise product analytics

Amplitude offers the deepest behavioural analysis here plus experimentation and feature flags in one platform, with a free tier of 2 million events and 10,000 replays. It publishes no rate above that; Vendr's contract data reports a median of $64,724 a year across 416 purchases as of February 2026. For a company evaluating it against GA360, that is roughly the same order of magnitude.

Should you leave GA4 at all

Often the answer is no, or not entirely. GA4 is free, it feeds Google Ads, and W3Techs puts Google Analytics on 47.2 percent of all websites in September 2026, which is 83.1 percent of sites whose analytics tool is known. Nobody should believe a claim that everyone is leaving.

Two more corrections worth making, because the internet repeats both. GA4 has not been ruled illegal in the European Union. The decisions people cite, from the Austrian, French, Italian, Swedish, Danish, Finnish and Norwegian regulators, all concerned pre-GA4 Google Analytics and turned on transfer law that predates the July 2023 EU-US Data Privacy Framework. The General Court dismissed the first challenge to that framework in September 2025 and it remains in force, but that ruling is under appeal at the Court of Justice, and the court assessed only the position as it stood when the framework was adopted. Consent obligations for analytics cookies are a separate question and still apply. And Universal Analytics data is gone: Google began permanent deletion in the week of 1 July 2024, so there is no historical archive waiting to be exported.

The sensible pattern for most stores is to keep GA4 for what it does free, add a tool that answers the commercial question, and stop asking GA4 for an attribution model it no longer has.

How to choose

Write down the question you are actually paying to answer. Which campaigns to cut points to attribution, and that is Arktis. How many people read the blog points to Plausible. Whether the data can leave your servers points to Matomo. Why users abandon step three points to Mixpanel, Amplitude or PostHog. Where visitors hesitate on the page points to Clarity, which is free.

To see attribution and recordings on your own traffic, start the 7-day Growth trial. Tiers are on the pricing page, and ecommerce attribution software compares ten tools built for the commercial question.

Sources

Google: GA4 data retention, the 2 and 14-month standard options and 360-only extensions

Google: about data sampling, 10 million events per ad-hoc query on standard properties

Google: about the (other) row, cardinality and undisclosed row limits

Google: BigQuery Export, the one-million-event daily cap, pause behaviour and streaming rate

Google: get started with attribution, the three remaining models and those removed in November 2023 and model data start dates

Google: behavioral modeling for consent mode, the 1,000-event and 1,000-user thresholds

Google: Universal Analytics has been replaced, permanent data deletion from July 2024

Cardinal Path: UA 360 vs GA4 360 pricing, the $50,000 suggested retail figure, August 2022

Plausible pricing page, accessed 16 September 2026

Matomo pricing page, Cloud and On-Premise, accessed 16 September 2026

PostHog pricing page, accessed 16 September 2026

Mixpanel pricing page, accessed 16 September 2026

Vendr: Amplitude contract data, median $64,724 across 416 purchases, February 2026

Didomi: European consent benchmark, 55.7 percent Western Europe opt-in, 2025 data

Backlinko, reporting GWI Q2 2025 data on ad blocker usage, 29.5 percent of internet users aged 16 to 64, page updated March 2026

Littledata: Shopify and Google Analytics tracking, February 2025, vendor research

W3Techs: Google Analytics usage, September 2026

IAPP: General Court dismisses the Latombe challenge to the EU-US Data Privacy Framework, September 2025, now under appeal at the Court of Justice