Google Analytics vs Matomo: when each option fits
GA4 fits teams whose measurement is tied to Google's stack and who have no one to run a server. If your paid search budget runs through Google Ads, your conversions feed Google bidding, or your reporting depends on Search Console and other Google-native links, GA4 is the path of least resistance. It also suits teams without hosting staff, because Google runs the collection, storage and interface. The trade-off is that you do not hold the raw data in a form you control: GA4's retention settings cap explorations, and Universal Analytics history is gone after Google's July 1, 2024 cutoff, so anything you need from GA4 should be exported on your own schedule. Confirm which Google integrations your campaigns rely on before you assume any replacement covers them.
Matomo On-Premise fits teams that need to own the raw visit data, keep it inside their own EU infrastructure and configure consent exemptions themselves. Matomo is GPLv3 and its On-Premise Community edition is free to run, so the recurring cost is hosting, database administration and upgrades, not licensing. Its CNIL-oriented configuration only works if you keep to the conditions in the regulator's guidance, and features such as heatmaps or session recordings are paid premium plugins that remove the exemption. Choose Matomo if you can staff a PHP and MySQL or MariaDB service, and if an imported GA4 history of aggregated reports is enough for continuity.
The decision turns on three questions: which integrations you cannot give up, which premium plugins you need, and who runs the server. Check whether Google Ads conversion import or Google-native reporting is a hard requirement. List any heatmap or session recording needs, since those add plugin licensing that falls outside the free core. Then name the owner for hosting, backups, upgrades and security patches. If that owner is unclear, GA4 or a managed Matomo Cloud plan may be the honest answer, and a fixed-scope implementation partner can scope the self-hosted build before you commit.
Which scenarios favor GA4 and which favor Matomo
This table maps common analytics scenarios to the tool that fits better and the condition that decides it. Read each row against your own traffic, team capacity and data rules, since the deciding condition is usually more important than the feature list. The main takeaway is that Matomo wins on data ownership and consent configuration, while GA4 keeps the edge where Google Ads and BigQuery are already part of the workflow.
| Scenario | Better fit | Condition that drives it |
|---|---|---|
| Google Ads conversion import | GA4 | Your bidding and conversion attribution depend on Google Ads linking, which Matomo does not replicate natively. Confirm what your ad accounts need before deciding. |
| BigQuery-based analysis | GA4 | If analysts already query GA4 data in BigQuery, the existing export and SQL models are a strong reason to stay. Matomo requires building your own export path from its database. |
| Small marketing team with no hosting capacity | GA4 | Matomo On-Premise means someone must operate the PHP application, database, backups and upgrades. If no one owns that work, the hosted Matomo Cloud option or GA4 may suit better. |
| French or EU site wanting analytics without a consent banner | Matomo | CNIL's exemption depends on strict configuration: audience measurement only, truncated IP addresses and a tracker lifetime of at most 13 months. Matomo can be set up to meet these conditions, while most large offerings do not qualify. |
| Heatmaps and session recordings | Neither as a default | Matomo's heatmap and session recording features are a paid premium plugin under the InnoCraft EULA, and adding them removes the consent exemption. GA4 does not provide these features either, so choose a dedicated tool if needed. |
| Regulated site with strict data location rules | Matomo | Self-hosted Matomo keeps raw visit data inside your own infrastructure, while GA4 data is processed by Google. Confirm where your data must reside and who can access the server. |
| Long-term raw data ownership and history | Matomo | GA4 retention settings cap explorations, and imported history in Matomo carries only aggregated reports, so the Visitor Log and segmentation will not cover imported periods. |
| Historic GA4 segment analysis after switching | GA4 | Segments and the Visitor Log do not work on data imported into Matomo. If you need raw-level analysis of past periods, export and retain GA4 data before cutover. |
Where your analytics data lives under each model
With GA4, Google hosts and operates the collection endpoint, the reporting interface and the stored event data. Your organization configures properties, data retention settings and any BigQuery export, but the underlying storage sits in Google's infrastructure and is governed by Google's terms. Reports are limited by the retention settings you choose, and anything not exported before the retention window closes is no longer available in the interface.
- Retention is set by Google's options: Retention settings cap explorations and event-level data in the interface, so you should confirm the current maximum for your property type in Google's documentation.
- Export depends on configuration: Data reaches your control only through exports such as BigQuery, which you must set up, pay for and secure yourself.
- Transfers follow the EU-US Data Privacy Framework: The European Commission adopted its adequacy decision on July 10, 2023 (decision 2023/1795), so personal data can flow to participating US companies under that framework. Confirm with your legal team whether Google's participation and your own transfer assessment satisfy your requirements.
- Identity and access are Google's platform: Access control runs through Google accounts and roles that you assign, but the platform's own operational logs and internal controls are not yours to inspect.
Check your exports, retention settings and transfer documentation before you rely on GA4 history for any decision or audit.
Matomo Cloud is a hosted service run by Matomo, with data stored in Frankfurt, Germany. Your organization gets a managed instance, so the vendor operates the servers, backups and upgrades, and you are the controller of the analytics data while Matomo acts as the hosting provider. The data location is in the EU, but the operating company and its subprocessors are still a third party, so the contract and processing terms decide what you can document.
- Location is fixed to Frankfurt: Matomo states that Cloud data is stored in Frankfurt, which helps with EU storage questions but does not remove the need to review subprocessors.
- Plans are tiered by hits: Cloud pricing depends on monthly hit volume and tier, so confirm the tier that matches your traffic before you commit.
- Operations stay with the vendor: Backups, patching and uptime targets are defined by the vendor's terms, so request the current service description in writing.
- Identity and logs are shared responsibility: You manage your users and roles in Matomo, while Matomo operates the underlying infrastructure and its audit trail.
- Plugins may be separate purchases: Premium features such as heatmaps are licensed separately, so check the cloud plan's included features.
Confirm the subprocessor list, the data processing terms and the Frankfurt storage commitment in the contract before you map Cloud into your records.
Matomo On-Premise runs on servers you control, whether in your own cloud account, a data center or a private network. The data never leaves that environment unless you send it elsewhere, so the location, the transfer analysis and the storage decisions sit with your organization. This also moves the operating work to your team: identity, access control, audit logging, backups and patching are yours to design and evidence.
- Data stays where you deploy it: Hit data, reports and database backups live in your chosen region and storage, which simplifies the transfer question for EU organizations.
- Access control is yours: You define administrator roles, user accounts, single sign-on integration and the rules for who can read raw visit data.
- Audit logging is yours to build: Web server, database and Matomo application logs must be retained and reviewed according to your policy, since nobody else holds them for you.
- Operations are a recurring cost: Upgrades such as the move to Matomo 6.0.0, which needs PHP 8.1 or newer and MySQL 8.0 or newer or MariaDB 10.6 or newer, must be planned and tested.
- Evidence is your responsibility: DeployOpen can prepare systems, controls and evidence for a compliance assessment, but independent auditors issue any report or certification.
Decide the hosting region, the log retention period and the access review cadence before you install, since these choices define your evidence set.
Can Matomo run without a consent banner in France?
France's CNIL allows some audience measurement to run without consent, but only when the tracker meets a tight set of conditions in Sheet 16. The purpose must be limited to audience measurement and A/B testing. The tracker must serve a single publisher scope, must not be cross-checked against other processing, must truncate the last byte of the IP address, and must keep its cookie lifetime to at most 13 months. Each condition is a design decision you own, not a checkbox the vendor ticks for you, so the first task is to write down how your Matomo instance meets each one.
CNIL also says that most large audience measurement offerings do not qualify for the exemption, regardless of how they are configured. Self-hosted Matomo is the open-source option the CNIL points to because you can configure it yourself, but the exemption is lost as soon as the setup goes beyond the conditions. Matomo's ePrivacy FAQ notes that adding a User ID, heatmaps or session recordings removes the exemption. If your marketing team wants Matomo's paid Heatmap and Session Recording plugin, plan for a consent banner on those pages before you install it.
Matomo's ePrivacy FAQ also names Italy, the Netherlands and Spain as authorities that exempt certain privacy-focused analytics from consent when criteria are met, and suggests following the CNIL configuration there. Confirm the rules with your privacy counsel and your own supervisory authority before you drop the banner, because the exemption applies to French law and other countries' rules may differ in detail. Document the configuration, the tracker lifetime and the list of excluded features so that an auditor or regulator can see the decision and when it was made.
How server size scales with monthly pageviews
Matomo publishes hardware floors by monthly pageviews, and its PHP and database requirements differ between Matomo 5 and the 6.0.0 beta. The cards below set out the server tiers, the runtime versions and the checks to make before you order hardware or plan an upgrade.
Single server up to 1 million pageviews
Matomo's requirements FAQ sizes one server for sites up to 1 million pageviews a month, with at least 4 CPUs, 8 GB of RAM and 250 GB of SSD. This tier suits a single marketing site or a small group of sites running the web server, PHP and database on one machine. Check storage first, because the database grows with every tracked hit. Plan disk for your retention period and report history, and confirm the figures against the FAQ before buying hardware.
Split servers up to 10 million
For sites up to 10 million pageviews a month, the FAQ recommends separate application and database servers, each with 8 CPUs and 16 GB of RAM. Splitting the two keeps report archiving and database queries from competing for the same memory and CPU. The cost is a second machine to patch, back up and monitor. Ask whether your traffic peaks sharply, because a monthly total hides the hours when load is highest.
Matomo 6.0.0 beta requirements
Matomo 6.0.0 needs PHP 8.1 or newer and either MySQL 8.0 or newer, or MariaDB 10.6 or newer. The changelog lists 6.0.0 as a beta, so treat it as a test target only. Do not run the beta on production analytics until a stable release is listed and your plugins and tracking setup have been tested against it. Confirm PHP and database versions in your operating system image before choosing a base build, since default packages on older distributions may be behind.
Matomo 5 runtime requirements
Matomo 4.16 and 5 run on PHP 8 with MySQL 5.5 or later, or MariaDB, behind a web server such as Apache, Nginx, IIS or LiteSpeed. The lower database floor means Matomo 5 is less demanding than 6.0.0 on the database side. Use the supported version list to choose the final stack, and confirm each PHP extension your plugins require. Check the requirements FAQ before you commit to a build, and test an upgrade path if you plan to move to 6.0.0 later.
Confirm the current stable release
Versions change faster than hardware plans, so confirm the latest stable release on the Matomo changelog before you size servers or write a change plan. When this page was prepared, the changelog listed 5.14.1 as the latest stable version and 6.0.0 in beta. Record the version, its PHP and database requirements, and the date you checked so the server build and the upgrade plan start from the same facts.
Plan for traffic beyond the tiers
The FAQ tiers stop at 10 million pageviews a month, so a larger property needs a sizing review with Matomo or an implementation partner before purchase. Ask how many sites share the instance, how much raw and report history you keep, and whether archiving finishes within your reporting window. Those answers change CPU and storage needs more than the pageview count alone. A property that currently pays for Google Analytics 360 may need a separate architecture, so ask for that in writing.
What moving from GA4 to Matomo involves
The sequence turns a GA4 property into a self-hosted Matomo site that holds its own raw visit data, with imported history kept as aggregated reports and both tools checked against each other before the Google tag is removed.
Size and provision the hosting
Your infrastructure team picks a hosting location inside your own cloud account or data center and sizes it against your monthly pageviews. Matomo's requirements FAQ gives a starting point: a single server with at least 4 CPUs, 8 GB of RAM and 250 GB of SSD for up to 1 million pageviews a month, and separate app and database servers with 8 CPUs and 16 GB of RAM each for up to 10 million. Confirm the PHP and database versions against the current requirements FAQ before you build. Done means a reachable Matomo host with a tested backup target and TLS certificate.
Install Matomo and the tracking code
Install Matomo On-Premise, create the website record and add the tracker through a direct script or your Tag Manager container. Decide early whether analytics will run under consent, because the CNIL conditions in France restrict which features and settings stay exempt. Your web team removes nothing from GA4 yet. Done means Matomo records test visits from staging and production, and the first pageviews appear in the Visitors reports.
Import Google Analytics history
Run the free Google Analytics Importer plugin to pull report data from the GA4 property into a new Matomo site. The importer brings over aggregated reports only, not raw visits and actions, so the Visitor Log, Ecommerce Log, Custom Reports and Segmentation will not work on the imported history, and some reports such as Funnels are not imported. Tell stakeholders this before they expect raw drill-downs. Done means the imported reports match a sample of GA4 figures for the same dates, within the differences you documented.
Rebuild goals and custom dimensions
Recreate the conversion goals, event definitions and custom dimensions that marketing uses, mapping each GA4 event name to a Matomo goal or custom dimension. Check that every tag you keep fires only the events you intend, and that no personal identifiers such as User ID are sent if you rely on the CNIL exemption. Done means a written mapping sheet signed off by the marketing owner, with each mapped metric tested on live traffic.
Run both tools in parallel
Keep the GA4 tag and the Matomo tracker running side by side for a period your analytics owner agrees. Compare visits, sources, conversions and key landing pages between the two tools, and record the differences that come from tracking method, consent rates and bot filtering. Done means a reconciliation report that explains each variance, approved by the people who use the numbers for budget decisions.
Cut over and retire the GA tag
Remove the GA4 tag from your Tag Manager container and website templates, then confirm that Matomo is the only analytics collector on the pages in scope. Update consent management settings and privacy notices to name Matomo as the processor. Keep the GA4 export and retention decisions documented, since GA4 data is not carried into Matomo at the raw level. Done means a tag audit showing no remaining Google Analytics requests and an updated privacy notice.
Plan the Google Ads and paid Matomo gaps
Decide which Google Ads conversion imports and audience links still need GA4, because Matomo does not replace those integrations. Confirm with your Google Ads account owner which links remain, and decide whether you need paid Matomo plugins such as the Heatmap and Session Recording plugin, which is sold under a separate license. Done means a written list of what stays on GA4, what moves to Matomo and which premium plugins your team has approved or rejected.
Questions buyers ask before moving from GA4 to Matomo
How much does a switch from GA4 to self-hosted Matomo cost?
The software itself carries no license cost: Matomo On-Premise Community is free, and its source is licensed under GPLv3. The spend sits in the components around it: servers or cloud instances sized by monthly pageviews, database and backup storage, the tracking and tag work on your sites, the one-time import and rebuild of goals and reports, ongoing upgrades and monitoring, and staff time. Recurring costs are mostly infrastructure and operations, while setup is one-time. Costs rise with pageview volume, number of sites, and any paid premium plugins. Ask for a written scope that lists each component separately, with who performs it, so you can compare it with your current GA4 spend, including any BigQuery export costs you carry today.
How long does a Matomo migration take, and what drives the timeline?
The duration depends mostly on how many sites, tags and custom reports you run, how clean your GA4 configuration is, and how many teams must sign off on dashboards. Server provisioning and the tracker change are usually the shortest steps; rebuilding goals, dimensions and reports, and running both tools in parallel before cutover take longer. Your internal capacity, change approval cycles and any consent or legal review also set the pace. No single duration applies to every organization, so ask the implementation partner for a timeline in the written scope, with milestones tied to the parallel-running period you choose.
Can our internal team run self-hosted Matomo without outside help?
Many teams can, provided they already operate PHP applications, MySQL or MariaDB, and a web server. Matomo's requirements FAQ lists PHP 8 and MySQL or MariaDB behind Apache, Nginx, IIS or LiteSpeed. Matomo 6.0.0 requires PHP 8.1 or newer and MySQL 8.0 or newer or MariaDB 10.6 or newer, so check your current versions first. Staff need ongoing ownership of upgrades, backups, security patches, archiving jobs and capacity planning. If you lack that skill set or want the work documented for audit evidence, you can engage DeployOpen for planning, implementation, hardening, backup and support under a fixed scope or a retainer.
Does Google Ads integration survive a move to Matomo?
No, not in the way GA4 provides it. Matomo can import aggregated report data from a Google Analytics property with its free Google Analytics Importer, but it does not replicate GA4's native link to Google Ads, audience sharing or conversion import. Teams that run paid search or display campaigns that depend on GA4 audiences or conversions should keep GA4 for those workflows, or run Matomo alongside it for on-site measurement. Confirm your Google Ads dependencies with your paid media team before deciding, and document which data each tool will be the source of.
What happens to our Universal Analytics history now?
Universal Analytics history is no longer available. Google states that standard Universal Analytics properties stopped processing data on July 1, 2023, and that access to Universal Analytics data, the interface and the API ended the week of July 1, 2024, after which unexported data was permanently deleted. Any history you still need must already exist in an export you made earlier. Your history from GA4 can be carried into Matomo only as aggregated report data through the importer, so raw visits, the Visitor Log and segmentation will not apply to imported periods. Check your archives now and confirm with your data owners which reports must be retained.
When is a premium Matomo plugin worth buying?
Premium plugins are worth evaluating only when a specific business need is unmet by the free core and by your own configuration. The Heatmap and Session Recording plugin is a paid premium plugin sold per number of users under the InnoCraft EULA rather than the GPL, and it offers a 30-day free trial. Buy it only after you have confirmed the licensing terms, the per-user pricing for your team size, and the consent impact. Matomo's ePrivacy FAQ states that heatmaps and session recordings remove the consent exemption, which can bring back a consent banner. Ask for the plugin cost in writing alongside any other component.
Will self-hosting Matomo remove the need for a consent banner in France?
Not automatically. CNIL recognizes some audience measurement trackers as exempt from consent when they meet conditions, including limiting purposes to audience measurement and A/B testing, truncating the last byte of the IP address, and keeping a tracker lifetime of at most 13 months. Most large audience measurement offerings do not fall within the exemption regardless of configuration, and CNIL suggests self-configurable open-source software such as Matomo. Adding User ID, heatmaps or session recordings removes the exemption. Have your legal or privacy team confirm your configuration against CNIL sheet 16 before relying on it.
How big a server does Matomo need for our pageviews?
Matomo's requirements FAQ sizes a single server with at least 4 CPUs, 8 GB of RAM and 250 GB of SSD for up to 1 million pageviews a month. For up to 10 million pageviews a month, it recommends separate app and database servers, each with 8 CPUs and 16 GB of RAM. Your real need also depends on archiving frequency, the number of sites, report complexity and retention length, so validate the sizing with a load test before cutover. Ask the implementation partner to state the sizing assumptions and the test results in the written scope.
When staying on GA4 is the right call
Keep GA4 if your paid media depends on it. Teams that run Google Ads conversion import, build audiences for remarketing, or rely on Google Tag Manager workflows tied to Google's own tags usually have little to gain from moving the whole measurement stack. Those integrations are built around GA4, and rebuilding them in Matomo means extra tagging work and a second source of conversion truth that someone must reconcile every month. Before deciding, list every campaign, audience and bidding strategy that reads GA4 data today, and confirm with your ads team which of them could be fed from Matomo or from server-side conversion uploads instead.
Switch when ownership or retention becomes the deciding requirement. A move makes sense when your legal or security team needs raw event data stored in infrastructure you control, when analysts need history longer than GA4's retention settings allow, or when EU control over where visitor data is processed has become a formal requirement. Matomo on your own servers answers all three, but it also means you take on hosting, backups, upgrades and the consent configuration that Google currently handles for you. Ask your analytics owner to write down which of these three requirements actually applies before committing to a cutover date.
A hybrid setup is a valid option to test. Many organizations run Matomo for owned website analysis, consent-exempt audience measurement where the CNIL conditions are met, and internal reporting, while keeping GA4 on paid landing pages for Google Ads. The cost is two tags on some pages and two numbers that will never match exactly, so the split only works if each tool has a named purpose in your reporting. Confirm with marketing and finance which KPIs must come from which tool before you add a second tag.
How DeployOpen helps evaluate and implement Matomo
DeployOpen is a Chainscore Labs company that plans, implements, hardens, backs up and documents self-hosted open-source platforms inside your own cloud accounts, data centers or air-gapped networks. For Google Analytics vs Matomo, we help you test the trade-offs in your environment, then build and hand over a Matomo On-Premise setup that you run and own.
How we scope
- Engagements are fixed-scope, with deliverables and timelines agreed in writing before work starts.
- We confirm which GA4 reports, goals and integrations, including Google Ads links, must survive the move.
- We size the Matomo servers against your actual monthly pageview volume and your database and backup needs.
How we build
- We install Matomo On-Premise in your environment and check it against the current PHP and MySQL or MariaDB requirements.
- We configure the tracker or Tag Manager and test whether consent-exempt settings fit your country's rules, or whether a consent banner is still needed.
- We import GA history through the Google Analytics Importer where it fits, and document which reports and raw-data views will not carry over.
How we hand over
- You receive written documentation of the architecture, configuration, sizing decisions and retention settings.
- We provide runbooks covering backups, upgrades, plugin management and the paid-plugin decisions you made during scoping.
- We run both tools in parallel for an agreed period so your team can compare reports before cutover.
How we support
- Ongoing support, upgrades and continuous testing are available as retainers after handover.
- Retainers can cover Matomo version upgrades, plugin updates and periodic checks of tracker configuration as your site changes.
- For compliance reviews, we prepare systems, controls and evidence, while independent auditors issue any assessment.
