Loading tutorials…
Loading tutorials…
Most GA4 → Plausible migrations fail because owners flip the switch on day 1 and lose historical context. The right pattern is a 30-60 day parallel run: both tools firing, both dashboards monitored, then a clean cutover. This is the playbook.
Who this is forOwners running GA4 today who are switching to Plausible for privacy/simplicity reasons. If you sell to EU customers and want to drop the cookie banner, or your team gave up on GA4's complexity, this is the migration path that does not lose data.
What you'll need
Step 1
GA4 will eventually delete your historical data. If you want it for posterity, export to BigQuery FIRST. Plausible cannot import GA4 history.
Plausible cannot import historical GA4 data — there is no migration path between the two data models.
If you want to preserve GA4 history for compliance, year-over-year analysis, or just peace of mind: set up GA4 → BigQuery export NOW. GA4 Admin → BigQuery Linking. This streams all events to BigQuery in near-realtime.
BigQuery storage costs ~$0.02/GB/mo. For a typical SMB doing 100K pageviews/month, that is <$5/mo for 2 years of history.
Without an export, the moment you delete the GA4 property (or 35 days after, whichever comes first), all history is gone permanently.
Alternative: download key reports manually from GA4 → Reports → Export. CSVs are clunky but free. Do this for: Top Pages, Acquisition by Channel, Conversions by Source/Medium, ecommerce purchases.
Step 2
Make a spreadsheet: GA4 Key Event Name → Plausible goal type + name. Every conversion you currently track must have a Plausible equivalent.
GA4 → Admin → Events. List every event marked as Key Event (formerly Conversions).
Open a spreadsheet. Three columns: GA4 Event Name | Plausible Goal Type | Plausible Goal Name.
For each GA4 Key Event, decide the Plausible mapping:
- GA4 'purchase' → Plausible revenue goal 'Purchase' with {revenue: {amount, currency}}
- GA4 'sign_up' → Plausible custom event goal 'Sign Up' (or pageview goal /welcome if you have a thank-you URL)
- GA4 'generate_lead' → Plausible custom event goal 'Lead' (or pageview goal /thank-you)
- GA4 'page_view' marked as Key Event → DO NOT replicate in Plausible (anti-pattern; see tutorial #3)
Document this mapping. Every team member should know the GA4 → Plausible equivalence by name.
Step 3
Both tools should fire for 30-60 days. Compare numbers daily for the first 2 weeks to validate the migration.
Install Plausible per tutorial #2. Do NOT uninstall GA4.
Both scripts should fire on every page. Yes, this means 2x the bytes — Plausible is ~1KB and GA4 is ~45KB, so the combined impact is still under 50KB.
Configure all Plausible goals to match the GA4 Key Event mapping from step 2. Fire events from the same codepaths.
For revenue events, ensure both GA4 (transaction_id + value + items array) and Plausible (revenue: {amount, currency}) fire from the same trigger.
Verify in Realtime that both tools register the same visit. GA4 Realtime + Plausible Realtime should each tick up by 1 within 60 seconds of your test visit.
Step 4
Every Monday, pull the previous week from both tools. Compare top metrics. Investigate gaps over 10%.
Schedule a recurring Monday morning task: pull previous-week data from both GA4 and Plausible.
Compare: total pageviews, unique visitors, top 10 pages, top 10 sources, top 5 conversion events (count and revenue).
Expect Plausible numbers to be LOWER than GA4 by 5-15% — Plausible's bot filter is more aggressive and it does not count any traffic where the user has disabled JS or blocked Plausible's script.
Larger gaps (>15%) mean something is misconfigured: missing pages, wrong data-domain, double-install, or an event not firing equivalently.
Document every discrepancy. By week 4, you should have <10% variance on all major metrics. By week 8, you should trust the Plausible numbers as your source of truth.
Step 5
Every spreadsheet, Notion doc, weekly report, and Looker Studio dashboard that reads from GA4 needs to be rewired or rebuilt.
List every place GA4 data is referenced: weekly KPI doc, monthly board deck, customer dashboards, Looker Studio embeds, Notion automations, Slack bot reports.
For each, decide: rebuild in Plausible (most can, with manual data entry or the Plausible Stats API) or accept that the report stops working post-migration.
Plausible has a public Stats API (Business plan) that can populate spreadsheets, Looker Studio (via community connectors), or custom dashboards.
For team reports, Plausible Business plan also supports CSV export of any view — Plausible CSV → import into your existing report templates.
Most teams find Plausible reports cleaner and DROP 30-50% of the GA4 reports they previously maintained. The simpler tool exposes how much GA4-specific cruft was accumulating.
Step 6
Tell your team the date when Plausible becomes the source of truth. Set a clear date — usually 30-60 days after install — and announce it.
Pick a cutover date — usually the start of a fiscal month for clean cohort math.
Announce internally: "From [date], Plausible is our source of truth for web analytics. GA4 will continue running for one more month as a fallback, then we remove it."
Update internal documentation: KPI definitions, source-of-truth references, weekly report templates — all should point at Plausible.
Train the team: 30-minute walkthrough of the Plausible dashboard. Show where the metrics they care about live.
Set a hard sunset date for GA4 (typically 30 days after the cutover). On that date, remove the GA4 script tag from the site.
Step 7
On the sunset date: remove the GA4 script, archive the property (do not delete — leave it for 35 days in trash), and remove the cookie banner if Plausible is your only tracker.
On sunset date, remove the GA4 script from your site templates. Deploy.
Verify in DevTools → Network that no requests go to www.google-analytics.com or analytics.google.com after deploy.
GA4 admin → Account Settings → Move property to trash. The property stays recoverable for 35 days. Do NOT permanently delete until you are 100% sure you do not need the historical reports.
If GA4 was your only tracker requiring consent: review your cookie banner. Plausible does not require consent (no cookies, no PII). You may be able to remove the cookie banner entirely.
Confirm with legal/privacy review before removing the banner. Other scripts (Meta Pixel, LinkedIn Insight, intercom chat) may still trigger consent requirements.
Celebrate: faster site, simpler analytics, no consent friction.
Common mistakes
Switching on day 1 with no parallel period
What goes wrong: You uninstall GA4 the same day you install Plausible. The Plausible numbers immediately look different from your historical GA4 data. Your team panics. Trust in analytics drops to zero. You spend 3 months trying to reconcile what happened.
How to avoid: Always run a 30-60 day parallel period. Both tools firing. Validate the variance. Only sunset GA4 once Plausible numbers are trusted.
Expecting Plausible numbers to match GA4 exactly
What goes wrong: Plausible shows 80K pageviews. GA4 shows 95K. Your team flags it as a bug. Actually it's normal — different bot filtering, different attribution windows, different exclusions. The variance is real but expected.
How to avoid: Expect 5-15% lower numbers in Plausible than GA4 due to stricter bot filtering and consent-rejection (GA4 still counts some consent-rejected visits as anonymous; Plausible does not). 5-15% variance is healthy.
Not exporting GA4 historical data
What goes wrong: Six months after the migration, you want to do year-over-year analysis. GA4 history was deleted when you removed the property. Plausible has no history before the migration. Year-over-year is impossible.
How to avoid: Before deleting the GA4 property, set up BigQuery export OR manually download key reports as CSV. ~$5/mo of BigQuery storage saves you the long-term loss.
Forgetting the Google Ads conversion tag
What goes wrong: You uninstall GA4. You forget that Google Ads imports conversions from GA4 to drive Smart Bidding. Smart Bidding loses its signal. Your CPA inflates 30-50% in the next 14 days.
How to avoid: Before removing GA4, install the Google Ads conversion tag directly (or via GTM). It is independent of GA4 and keeps Smart Bidding fed even after GA4 is gone. See tutorial #6.
Mapping page_view as a Key Event from GA4 to Plausible
What goes wrong: Your GA4 had page_view marked as a Key Event (anti-pattern, see ga4 tutorials). You replicate this in Plausible by marking every pageview as a goal. Now you have noisy goals, billing inflation, and unusable reports.
How to avoid: Do NOT replicate the page_view-as-Key-Event mistake. Use the migration to fix the goal taxonomy: only real outcomes (purchase, signup, lead, demo) are goals. Pageviews are tracked but not flagged as conversions.
Not removing the cookie banner after migration
What goes wrong: You finish the migration. Plausible is your only tracker. You forget to update the cookie banner config. Your site still shows the consent prompt — which now does nothing because Plausible does not need consent. Bad UX for no reason.
How to avoid: After the migration, audit all scripts. If Plausible is your only consent-requiring tracker (and it is not consent-requiring), check whether you can remove the cookie banner. Confirm with legal review first.
Recap
Done — what's next
How to set up a Plausible Analytics account the right way
Read the next tutorial
Hand it off
GA4 to Plausible migration is the analytics equivalent of moving house. The packing (goal mapping) takes longer than the moving day. A vetted specialist who has migrated 10+ sites can run the full 60-day playbook with you — typically 30-40 hours over 2 months, $400-600 total at $14-16/hr — and your team trusts the new numbers from week 4 instead of month 6.
See specialist rates
No. The two have fundamentally different data models. GA4 is event-based with granular parameters; Plausible is pageview-based with limited dimensions. Historical GA4 data can be exported to BigQuery for posterity but cannot be replayed into Plausible.
30 days minimum, 60 days recommended. Longer for high-stakes accounts (e-commerce above $100K/mo revenue, ad spend above $20K/mo). The first 14 days catch obvious gaps; days 30-60 surface subtle attribution drift.
Generally yes for last-non-direct click attribution (Plausible default). For first-click or multi-touch comparisons, no — Plausible only supports last-non-direct. If your team uses GA4's data-driven attribution, Plausible is a downgrade in attribution depth.
Yes — many teams do, especially when ad attribution requires GA4 → Google Ads import. Plausible is the human-friendly dashboard; GA4 is the ad-platform-integration layer. The combined cost is GA4 (free) + Plausible ($9-39/mo for most), with negligible script-weight impact.
Small site (single page, no SaaS app, <100K pageviews/mo): 1-2 weeks. Medium SaaS (multiple page templates, 5-10 goals, custom events): 4-6 weeks including parallel run. Enterprise with multi-domain + custom dashboards: 8-12 weeks. Plan for the parallel-run period, not the install time.
Plausible's reports are intentionally simpler — that is the value proposition for owners who never used Explorations effectively. If your team relies on GA4 Explorations or BigQuery analyses, Plausible is a downgrade. Consider PostHog instead, which has analytics depth closer to GA4 with the privacy posture closer to Plausible.
Plausible Analytics
Plausible is the easiest analytics tool to install — which is exactly why owners breeze through setup and lock in choices they later regret. The plan tier, timezone, and shared-link defaults are all hard to fix once data is flowing.
Plausible Analytics
Plausible's goal system is brutally simple — and that simplicity is what makes it powerful. Three types: pageview goals, custom events, and revenue goals. Most owners set up the wrong type, fire too many events, and blow past their plan. This is the discipline that prevents both.
Plausible Analytics
All three are credible GA4 alternatives — privacy-first, no cookies, lightweight scripts. The right one depends on your feature needs, where your team operates, and how aggressively you need to optimize on price. This walks through the honest tradeoffs.
Google Analytics 4
DIY GA4 is a great idea — until it isn't. This is the honest framework: when the cost of unreliable data exceeds the cost of hiring help, and how to tell which side you're on.