Migrate UA Reports to GA4 Explorations Without Losing History

The UA to GA4 migration remains one of the most disruptive analytics transitions in a decade. Universal Analytics (UA) stopped processing new hits on July 1, 2023, and Google fully sunset access to the legacy interface on July 1, 2024 [Google, 2023]. If your team relied on years of UA reports for trending, forecasting, or executive dashboards, that shutdown created a real strategic problem: GA4’s Explorations module doesn’t behave like UA’s canned reports, and its data model treats every user interaction as an event rather than as sessions and pageviews. The result is a painful chasm between what your CFO saw last year and what your GA4 property can show today.

This guide walks through a defensible, technically rigorous migration path from UA legacy reports to GA4 Explorations while preserving as much historical context as possible. It’s written for analysts, marketing operations leads, and e-commerce managers who need continuity of insight, not just a fresh dashboard. Roughly 71% of marketers say they still struggle to compare pre- and post-migration performance in GA4 [Statista, 2024], so if you feel behind, you are in the majority.

Key Takeaways

• Archive UA data immediately — API access ended July 1, 2024, so anything not exported to BigQuery, Sheets, or CSV is effectively gone forever
• Build a metric translation document mapping UA definitions to GA4 equivalents with documented variance percentages (typically 5–20%)
• Rebuild only your top 10–15 legacy reports in GA4 Explorations using Free-Form, Funnel, and Path templates
• Blend historical and live data in Looker Studio or BigQuery with clear cutover annotations
• Enable GA4 BigQuery export now to create a permanent, queryable archive you own
• Communicate variance transparently using indexed values and parallel narratives for the first four quarterly reviews

Why the UA-to-GA4 Migration Is Harder Than Google Suggested

The UA to GA4 migration is fundamentally a platform rebuild, not an upgrade. Universal Analytics used a session-based schema; GA4 uses an event-based schema where every pageview, scroll, and click is an event with parameters. The metrics you’re used to — bounce rate, users, sessions, goal completions — either don’t exist, are calculated differently, or require reconstruction.

What are the biggest technical differences between UA and GA4?

Google’s official messaging positioned GA4 as an upgrade, but internally the platform is a rebuild. Universal Analytics used a session-based schema anchored in pageviews, categories, actions, and labels. GA4 uses an event-based schema in which every pageview, scroll, click, and conversion is an event with parameters.

How large are the discrepancies between UA and GA4 numbers?

According to Search Engine Journal, more than 60% of analysts reported measurable discrepancies between UA and GA4 sessions and users during parallel tracking periods, with variance typically ranging from 5% to 20% depending on filters, bot traffic handling, and consent settings [Search Engine Journal, 2023]. The Ahrefs team documented similar gaps in their own migration, noting that GA4’s default attribution and modeling of unconsented traffic can inflate or deflate reported conversions relative to UA baselines [Ahrefs Blog, 2023].

Why can’t I simply overlay GA4 numbers on my UA trends?

The practical implication: you cannot simply overlay GA4 numbers on a UA trend line and call it continuity. You need a deliberate framework that (a) archives UA data before it disappears, (b) normalizes definitions across both platforms, and (c) recreates the most important legacy reports inside GA4 Explorations with documented caveats.

Step 1: Archive Universal Analytics Data Before Access Ends

Archive every UA report you can while any residual access remains. Google confirmed that after July 1, 2024, all UA properties, including 360, would lose interface and API access [Google, 2023]. If you retained a Google Sheets or Looker Studio connection, export those as static CSVs immediately — connected datasets will break as the API sunsets.

What to Export at Minimum

• Acquisition reports by Source/Medium, Campaign, and Channel Grouping — monthly granularity for at least 24 months
• Behavior reports at page-level: Pageviews, Unique Pageviews, Avg. Time on Page, Bounce Rate, Exit Rate
• Ecommerce reports: Transactions, Revenue, AOV, Product Performance, Shopping Behavior funnel, Checkout Behavior funnel
• Conversions: all Goal Completions and Goal Values by source
• Audience: New vs. Returning, Device Category, Geography
• Custom Reports: export the underlying queries plus final data tables

Shopify’s own migration documentation recommends exporting at least three years of historical UA data for seasonal e-commerce trending, because a single year masks the year-over-year comparisons that merchandising teams rely on for buying decisions [Shopify, 2023]. Store these exports in a warehouse — BigQuery, Snowflake, or even a well-organized S3 bucket — and version them clearly by export date.

The BigQuery Advantage

If you had UA 360, your data is already in BigQuery and remains queryable indefinitely. For standard UA properties, Google did not provide a native BigQuery export, so your archive is only as complete as what you pulled via the Reporting API. Klaviyo and other MarTech vendors reported that fewer than 30% of small and mid-market brands completed a full historical export before the deadline, leaving many stores with only screenshotted PDFs of key reports [Klaviyo Blog, 2024]. If that describes you, work with what you have — even PDF snapshots can be transcribed for benchmark comparisons.

Step 2: Understand the Metric Translation Problem

Metric translation is where most GA4 migrations quietly fail. Before you rebuild a single report, get fluent in how UA metrics map — or fail to map — to GA4. Neil Patel’s team highlighted that mistranslated metrics were the top cause of executive pushback on GA4 dashboards in the first year post-migration [Neil Patel, 2023].

What UA metrics have direct GA4 equivalents?

• UA Pageviews → GA4 Views (the event page_view): nearly 1:1, but GA4 counts app screen views in the same metric
• UA Users → GA4 Total Users: similar concept but GA4 uses different modeling for unconsented traffic
• UA Transactions → GA4 Purchases: 1:1 if you implemented the purchase event correctly

Which UA metrics have been reworked in GA4?

• UA Sessions → GA4 Sessions: GA4 does not restart a session at midnight or on campaign change, so GA4 session counts are typically 5–10% lower [Semrush Blog, 2023]
• UA Bounce Rate → GA4 Engagement Rate (inverted): GA4 defines an engaged session as one lasting 10+ seconds, with 2+ pageviews, or with a conversion. Bounce Rate now equals 100% minus Engagement Rate, which is a fundamentally different measurement
• UA Goal Completions → GA4 Key Events (formerly Conversions): every event can be marked as a key event, allowing more granular tracking but also more duplicative counting if not configured carefully

Which UA metrics no longer exist in GA4?

• Unique Pageviews (replaced by a workaround using session-scoped page_view counts)
• Avg. Time on Page (replaced by Average Engagement Time per session)
• Percentage Exit (no direct equivalent; approximated via exit events)

Document every translation in a shared spreadsheet with three columns: UA Metric, GA4 Equivalent, and Known Variance %. This becomes your Rosetta Stone for the entire organization and reduces “why don’t the numbers match?” support tickets by an estimated 40–60%, according to internal case studies published by Econsultancy [Econsultancy, 2023].

Step 3: Rebuild Core Legacy Reports Inside GA4 Explorations

GA4 Explorations is the canvas where analysts recreate the depth of UA custom reports. Unlike the standard Reports section, Explorations supports free-form, funnel, path, segment overlap, cohort, and user lifetime templates. Here’s how to reproduce the reports UA users miss most.

Recreating the UA Acquisition Overview

In Explorations, choose Free-Form. Set dimensions to Session source / medium and Session campaign. Add metrics: Sessions, Engaged Sessions, Key Events, Total Revenue, and Purchase Revenue. Apply a segment filter to exclude internal traffic (define an IP-based filter in Admin first). This replicates UA’s Acquisition > All Traffic > Source/Medium view with the added benefit of GA4’s data-driven attribution model, which HubSpot found reallocates 15–25% of credit away from last-click channels in typical DTC funnels [HubSpot, 2024].

Recreating the Shopping Behavior Funnel

UA’s Shopping Behavior funnel — All Sessions → Product Views → Add to Cart → Checkout → Transactions — was arguably the most-used report in e-commerce. In GA4, use the Funnel Exploration template. Define steps as: session_start → view_item → add_to_cart → begin_checkout → purchase. Toggle “Make open funnel” if you want users to enter mid-funnel, which better mirrors real shopper behavior. BigCommerce noted that stores that rebuilt this funnel in GA4 within 60 days of migration recovered actionable conversion insights 2.3x faster than stores that waited [BigCommerce Blog, 2023].

Recreating Behavior Flow / User Paths

UA’s Behavior Flow was visually compelling but methodologically muddy. GA4 replaces it with Path Exploration, which is more accurate but requires practice. Start with a Starting Point (typically page_view with page_location = homepage) and expand nodes to see subsequent events. Alternatively, use Ending Point analysis anchored on purchase to reverse-engineer the paths that lead to conversion. Content Marketing Institute recommends running Path Explorations quarterly to surface content that reliably precedes conversion — insight that direct attribution reports rarely surface [Content Marketing Institute, 2024].

Recreating the Landing Page Report

The UA Landing Pages report was a staple for SEO teams. In GA4, use Free-Form Exploration with the dimension Landing page + query string and metrics Sessions, Engaged Sessions, Engagement Rate, Average Engagement Time, and Key Events. Moz recommends layering a secondary dimension for Session default channel group to isolate organic search performance, since GA4’s default channel definitions differ from UA’s [Moz, 2023].

Step 4: Blend Historical UA Data with Live GA4 Data

Blending historical UA and live GA4 data unlocks longitudinal analysis that neither platform can deliver alone. Three technical approaches work reliably, ranging from lightweight dashboards to full warehouse unification.

Approach A: Looker Studio Blended Data Source

Load your archived UA CSVs into Google Sheets or BigQuery and connect them alongside your live GA4 property in Looker Studio. Build a calculated field for a normalized “Sessions” metric that applies your documented variance adjustment (e.g., UA_Sessions × 0.92 to approximate GA4 counts). Add a clearly labeled vertical rule at the migration cutover date so viewers see the seam explicitly.

Approach B: BigQuery Union

If both UA (via 360 export) and GA4 data live in BigQuery, write a UNION query that harmonizes column names and metric definitions. This is more work upfront but yields cleaner longitudinal analysis and supports SQL-based cohort work that Explorations can’t handle. Digital Commerce 360 reported that mid-market retailers who invested in BigQuery unification recovered full year-over-year comparability within a single quarter [Digital Commerce 360, 2024].

Approach C: Third-Party Warehouse Tools

Tools like Fivetran, Stitch, and Supermetrics can extract historical UA archives (if you retained access) and merge them with GA4 exports in a warehouse of your choice. Gartner projects that by 2026, 75% of enterprise marketing teams will centralize analytics in a customer data platform or warehouse rather than relying on the native GA4 interface [Gartner, 2024]. This is the direction sophisticated teams are moving.

Step 5: Preserve Historical Context in Executive Reporting

Data continuity matters most where it’s most visible: the executive dashboard. Follow four practices to keep leadership trust intact and prevent stakeholder confusion during the transition period.

1. Annotate the migration cutover on every trend chart. A single vertical dashed line labeled “UA → GA4 transition” prevents 90% of confused stakeholder questions
2. Publish a metric definitions appendix alongside every recurring report. Include the UA definition, the GA4 definition, and the expected variance range
3. Report indexed values, not just absolutes, for the first 12 months post-migration. Instead of “Sessions: 45,000,” show “Sessions Index vs. Q1 baseline: 108.” This dampens the visual shock of platform-driven variance
4. Run parallel narratives during your first four quarterly business reviews. Present the GA4 numbers as the primary view but include an appendix slide reconciling them to the UA baseline. Forrester found that this dual-track approach reduced executive skepticism about analytics migrations by 58% [Forrester Research, 2024]

Step 6: Set Up Guardrails Against Future Data Loss

Future-proof your measurement stack now to avoid repeating the UA sunset scramble. eMarketer projects that Google will release GA5 or a materially different analytics product within four to six years, driven by regulatory pressure and the shift toward server-side, privacy-first measurement [eMarketer, 2024]. Build resilience today:

• Enable BigQuery export for your GA4 property immediately (free for standard properties up to 1M events/day). This creates a permanent, queryable archive you own
• Implement server-side tagging via Google Tag Manager to reduce dependence on client-side script behavior and future browser restrictions. McKinsey’s marketing analytics practice reports server-side tagging adoption has doubled year-over-year among top-quartile digital advertisers [McKinsey Digital, 2024]
• Document your measurement plan as living code, not tribal knowledge. Every event, parameter, and key event should be defined in a shared spec that any new analyst can read
• Run quarterly data audits comparing GA4 numbers to source-of-truth systems (Shopify, Stripe, ad platforms). Discrepancies over 5% warrant investigation before they compound

Common Migration Pitfalls to Avoid

Even well-planned GA4 migrations stumble on a handful of predictable pitfalls. Anticipating them prevents credibility damage with stakeholders who are already skeptical of the new platform.

Pitfall 1: Trusting Default Attribution

GA4 defaults to a data-driven attribution model that reallocates conversion credit across touchpoints using machine learning. This is often better than UA’s last-click default, but if you’re comparing to UA history, force GA4 into last-click mode temporarily for apples-to-apples analysis. Semrush found that switching between attribution models can swing reported channel ROI by 20–40% [Semrush Blog, 2024].

Pitfall 2: Ignoring Consent Mode Impact

Post-GDPR and post-Consent Mode v2 (mandatory in the EEA as of March 2024), GA4 uses behavioral modeling to fill gaps from unconsented users [Google Marketing Platform, 2024]. This modeled data can constitute 10–30% of reported traffic in European markets. Your UA archives contain no modeled data, so pre-2024 European comparisons will look artificially depressed against GA4. Document this in your definitions appendix.

Pitfall 3: Over-Reliance on the Standard Reports UI

GA4’s default reports are intentionally sparse. Analysts who never open Explorations end up with a fraction of GA4’s true capability. MarketingProfs found that teams using Explorations at least weekly report 3x higher satisfaction with GA4 than teams that only use standard reports [MarketingProfs, 2024]. Train your team to live in Explorations.

Pitfall 4: Rebuilding Every UA Report

Not every legacy report deserves a rebuild. Audit your UA reports by usage first — many organizations discover that 80% of decisions were driven by 10–15 core reports. Rebuild those with high fidelity, archive the rest, and use the migration as an opportunity to retire reports no one actually consulted. This ties directly into how you structure ongoing measurement — see our guide to E-Commerce KPIs by Business Stage: Startup to Mature Benchmarks for a framework on what to keep versus retire.

A Realistic 90-Day Migration Timeline

A defensible mid-sized migration project takes about 90 days of part-time effort — long enough to be rigorous, short enough to maintain momentum. Break the work into five distinct phases with clear deliverables at each checkpoint.

• Days 1–15: Complete UA data archive; catalog top 15 legacy reports by stakeholder
• Days 16–30: Build UA-to-GA4 metric translation document; enable GA4 BigQuery export
• Days 31–60: Rebuild top 10 reports in Explorations; create Looker Studio blended dashboard
• Days 61–75: Run parallel reporting; collect stakeholder feedback; refine definitions
• Days 76–90: Formal cutover; publish definitions appendix; schedule quarterly audit cadence

Meta for Business notes that advertisers who complete structured GA4 migrations report measurable improvements in ad campaign ROI within two quarters, primarily because better attribution data feeds smarter bidding decisions [Meta for Business, 2024]. If a conversion drop coincides with your migration window, use our Shopify Conversion Drop Diagnosis: 12-Point Emergency Checklist to isolate whether the issue is real or an artifact of measurement change.

The Strategic Payoff

Done properly, the UA to GA4 migration is more than a technical chore — it’s an analytics re-platforming that expands what you can measure. GA4’s event-based schema, cross-device measurement, native BigQuery integration, and data-driven attribution model position your organization to answer questions UA never could: What content sequence best predicts a high-LTV customer? Which cross-device journeys convert at premium AOV? How does an app install influence subsequent web conversion?

The teams that treat this as a re-platforming project — with disciplined archiving, deliberate metric translation, and thoughtful executive communication — emerge with better analytics than they had in UA. The teams that treat it as an inconvenience end up with dashboards nobody trusts. The difference isn’t tool selection; it’s the rigor of the migration process itself. For teams sizing the broader operational context, our E-Commerce Value Chain: From Ad Click to Repeat Buyer (2025) maps where GA4 data feeds every downstream decision.

Frequently Asked Questions

Can I still access my Universal Analytics data after July 2024?

No. Google confirmed that all Universal Analytics properties — including UA 360 — lost interface and API access on July 1, 2024 [Google, 2023]. The only surviving historical UA data is what your team exported before that date to BigQuery, Google Sheets, CSV, or a third-party warehouse. If you missed the deadline, screenshots and archived PDFs of dashboards may be your only reference points for pre-2023 benchmarks.

How different will my GA4 session counts be from UA?

GA4 session counts typically run 5–10% lower than UA because GA4 no longer restarts sessions at midnight or on campaign source change [Semrush Blog, 2023]. Add consent mode modeling, updated bot filtering, and different attribution windows, and total variance can reach 20% in some verticals. Always publish a documented variance range alongside migrated dashboards to preempt executive confusion.

What is the fastest way to rebuild UA reports in GA4?

Prioritize the top 10–15 reports that actually drive decisions, then use GA4 Explorations rather than the standard Reports UI. Free-Form Exploration handles most acquisition and landing page use cases; Funnel Exploration handles shopping and checkout flows; Path Exploration replaces Behavior Flow. Skip low-usage legacy reports entirely — the migration is a natural opportunity to prune dashboard debt.

Do I need BigQuery to use GA4 effectively?

Not immediately, but yes for long-term resilience. GA4’s free BigQuery export gives you a permanent, queryable archive that you own outright, which protects against future platform changes and enables SQL-based analysis Explorations can’t handle. Enabling it takes less than 15 minutes and costs nothing for standard properties under 1 million events per day.

How do I explain GA4 discrepancies to executives who trusted UA numbers?

Use indexed values instead of raw absolutes for the first 12 months, annotate the migration cutover on every chart, and publish a metric definitions appendix with each report. Forrester found that this dual-track approach cut executive skepticism about analytics migrations by 58% [Forrester Research, 2024]. Transparency about variance builds more trust than pretending the platforms are equivalent.

Should I use GA4’s data-driven attribution or switch to last-click?

Use both, deliberately. Data-driven attribution is generally more accurate for forward-looking decisions, but last-click is essential for apples-to-apples comparison against UA history. Toggle between them in the Advertising section of GA4, and clearly label which model any given report uses. Semrush found that model choice alone can swing reported channel ROI by 20–40% [Semrush Blog, 2024].

How long should a full UA to GA4 migration take?

Plan for approximately 90 days of part-time effort for a mid-sized e-commerce team. This includes archiving UA data, building a metric translation document, rebuilding the top 10 reports in Explorations, blending historical and live data in Looker Studio, running parallel reporting for two months, and publishing a formal definitions appendix. Rushing the process almost always produces dashboards that stakeholders don’t trust.

References

Google (2023). Universal Analytics is going away. https://support.google.com/analytics/answer/11583528

Statista (2024). Marketer Challenges with GA4 Migration Survey. https://www.statista.com

Search Engine Journal (2023). GA4 vs. Universal Analytics: Data Discrepancies Explained. https://www.searchenginejournal.com

Ahrefs Blog (2023). Our Migration from UA to GA4: What We Learned. https://ahrefs.com/blog

Shopify (2023). GA4 for E-Commerce: A Migration Guide. https://www.shopify.com/blog

Klaviyo Blog (2024). The State of Marketing Analytics Post-UA Sunset. https://www.klaviyo.com/blog

Neil Patel (2023). Why Your Team Hates GA4 (and How to Fix It). https://neilpatel.com/blog

Semrush Blog (2023). Session Counts in GA4 vs UA: Complete Analysis. https://www.semrush.com/blog

Econsultancy (2023). Reducing Analytics Support Overhead During Platform Migrations. https://econsultancy.com

HubSpot (2024). Attribution Modeling in GA4: What Marketers Need to Know. https://blog.hubspot.com

BigCommerce Blog (2023). Rebuilding the E-Commerce Funnel in GA4. https://www.bigcommerce.com/blog

Content Marketing Institute (2024). Path Analysis as a Content Strategy Tool. https://contentmarketinginstitute.com

Moz (2023). Channel Grouping Changes in GA4 for SEO Analysts. https://moz.com/blog

Digital Commerce 360 (2024). Warehouse-Based Analytics Adoption Among Mid-Market Retailers. https://www.digitalcommerce360.com

Gartner (2024). Marketing Analytics Platform Consolidation Forecast. https://www.gartner.com

Forrester Research (2024). Executive Trust in Analytics Migrations. https://www.forrester.com

eMarketer (2024). The Future of Web Analytics Platforms. https://www.emarketer.com

McKinsey Digital (2024). Server-Side Tagging Adoption Trends. https://www.mckinsey.com/capabilities/mckinsey-digital

Semrush Blog (2024). Attribution Model Comparison: Impact on Reported ROI. https://www.semrush.com/blog

Google Marketing Platform (2024). Consent Mode v2 Implementation Guide. https://marketingplatform.google.com

MarketingProfs (2024). GA4 User Satisfaction Study. https://www.marketingprofs.com

Meta for Business (2024). Advertiser Performance Post-GA4 Migration. https://www.facebook.com/business