Migrate From Google Analytics 4: A Step-by-Step Plan

To migrate from Google Analytics 4, audit the reports you use, install the new tool next to GA4, recreate key events as goals, compare both for several weeks, export history you need, and only then remove the GA4 tag. Parallel running is what makes the switch safe.

Step 1: Audit what you actually use

Open GA4 and list the reports opened in the last 90 days, who opens them and what decision each informs. Add every conversion event, every audience used in Google Ads and every scheduled email report. This inventory is your migration scope. Anything not on it, you can leave behind without regret.

Include integrations. Is Search Console linked? Do ads import conversions from GA4? Does a dashboard pull data through an API? Each of these needs a replacement or a deliberate decision to drop it.

Step 2: Choose and set up the new tool

Choosing is covered in privacy-friendly analytics tools. For the example in this guide we use BusinessMCP, whose web analytics page describes a free plan for up to 300 visitors a month, so you can test without a card. Whatever you choose, check retention, export options and the DPA before you start.

Step 3: Install alongside GA4

Add the new script to the <head> of every page. BusinessMCP’s install documentation says one <script defer> tag with your tracker ID is the single required step, with platform notes for Google Tag Manager, WordPress, Shopify, Webflow, Framer, Next.js, Nuxt, Astro, SvelteKit and plain HTML. Use only one tag per page for a given tool, since two copies double-count. Our script install guide goes through verification.

Step 4: Recreate key events as goals

GA4 item	Replacement action	Check
Key events (conversions)	Create a named goal for each	Fire it once on a test page
Purchase event	Record revenue with a numeric value or connect your payment provider	Compare with your order records
Form submit	Track on the thank-you action, not the button click	Submit a test entry
Campaign UTMs	Keep the same UTM scheme	Open a tagged link, confirm the source
Audiences for ads	Decide: keep GA4 for ads, or rebuild elsewhere	Confirm with the ads owner

Key takeaway: migrate goals first. A tool with traffic but no conversions will not convince anyone to switch off GA4.

Keep your UTM rules consistent. BusinessMCP’s UTM parameters guide is a reasonable reference for naming conventions you can apply in any tool.

Step 5: Compare for two to four weeks

Do not expect matching totals. GA4 uses cookie-based users and sessions, while cookieless tools count visitors by their own rules. Compare:

1. Rank order of top ten pages.
2. Rank order of top channels.
3. Week-over-week direction of traffic.
4. Conversion counts against your real orders or leads.

If the order and direction agree and conversions line up with your own records, the new tool is giving you the signal you need. If conversions are missing, debug before moving on.

Step 6: Bring history across

You have two options. Export GA4 data you want to keep (reports as CSV, or via BigQuery if enabled) and store it. Or use an importer if the new tool has one. BusinessMCP imports GA4 history at day level (users, sessions, conversions) and displays it as a labelled “before you had us” series, never merged into first-party totals, as the comparison page explains. That separation is honest: the definitions differ, so a combined chart would mislead.

Step 7: Switch off GA4 carefully

1. Tell stakeholders the date and where to find each report.
2. Remove the GA4 tag, or keep it for ads only if needed.
3. Keep read access to the GA4 property for a few months.
4. Update your privacy notice and cookie settings to reflect the new setup. See GDPR-compliant web analytics.

If you keep Google tags for ads, read Google Consent Mode v2 explained since those tags still need consent handling.

Common pitfalls

• Switching before goals work. You lose a month of conversion data.
• Forgetting subdomains. Check checkout and booking subdomains.
• Missing single-page app navigation. Verify route changes count as pageviews.
• Skipping the privacy notice. Update it the same day you switch.

A realistic timeline

A small site can finish in one afternoon plus a month of parallel running. A larger site with ads, multiple domains and stakeholders should plan a quarter. The work is mostly communication, not code.

What to tell stakeholders

A migration succeeds when people trust the new numbers. Send a short note before you switch that covers three things: the date, where to find each familiar report, and why totals will differ. A sentence such as “unique visitors are counted by a different rule, so expect the new figure to be lower or higher, but trends and rankings should agree” prevents most of the awkward questions. Attach a one-page mapping from old reports to new ones.

Keep the parallel period honest. If the new tool disagrees on direction, for example traffic up in one and down in the other, investigate before moving on. Common causes are a missing tag on some pages, a bot filter difference, or a goal firing twice.

When to keep GA4 a little longer

If your ads team depends on GA4 audiences or conversion import, keep the tag until you have a replacement workflow. You can retire the reporting use of GA4 while leaving it connected for ads, as long as the consent setup for Google’s tags is correct; see Google Consent Mode v2 explained. For a fuller view of what you gain and give up, read the Google Analytics alternative comparison from the vendor, alongside independent sources, and the alternatives overview here.

After the switch: the first review

Schedule a review thirty days after GA4 goes quiet. Check that goals still fire, that campaign links are still tagged, and that the report people use most is recreated. Ask each stakeholder one question: what do you miss? Their answers are your backlog. Some will be fixable with a saved view or a scheduled export, some will need a second tool, and some will turn out to be reports nobody needed.

Archive the migration notes next to the privacy record: what was installed, when, which events map to which goals, which data was exported and where it is stored. When someone asks in a year why the numbers jumped in a given month, that single page explains it. It also makes the next tool change faster, because most of the mapping work is already done.