GA4 vs Universal Analytics: The Ecommerce Setup Guide for 2026

If you are running an ecommerce store in 2026 and your analytics still feels like guesswork, you are burning between 15% and 30% of your marketing budget on channels you cannot actually measure. Universal Analytics has been sunset for over two years, the reporting data is gone, and yet we still open client accounts every month where the migration was done badly or never finished. This is the head-to-head you needed back in 2023, written for the reality of running a store now: what Google Analytics 4 actually does, what Universal Analytics did that people miss, and exactly how to set GA4 up so it earns its keep.

TL;DR

• GA4 Is The Only Live Option: Universal Analytics stopped processing data on July 1, 2023, and its interface was fully retired in 2024, so a GA4 vs UA comparison in 2026 is really about understanding what changed and configuring GA4 correctly rather than choosing between them.
• The Event Model Is The Real Difference: GA4 measures everything as events with parameters instead of sessions and pageviews, which is more flexible and future-proof but demands a deliberate setup, especially for ecommerce purchase and funnel tracking.
• Setup Quality Determines Data Quality: In our audits, roughly 40% of GA4 properties are missing enhanced ecommerce events, key events, or a clean data stream, so a correct Google Analytics 4 setup tutorial matters more than the platform debate itself.

Why This Comparison Still Matters In 2026

Let us clear something up first, because it saves confusion later. Universal Analytics is not a live product you can choose today. Google stopped processing new hits in standard UA properties on July 1, 2023, and the historical interface was decommissioned in 2024. So when store owners ask us whether they should run GA4 or Universal Analytics, the honest answer is that the decision was made for you.

But the comparison is far from academic. Most operators built their mental model of web analytics on Universal Analytics: sessions, bounce rate, goals, ecommerce plugins that just worked. GA4 speaks a different language, and if you carry the old assumptions into the new tool, your reports will look wrong, your numbers will not reconcile with your ad platforms, and you will make bad decisions with confident-looking dashboards.

At Presta, we’ve migrated and rebuilt analytics for dozens of ecommerce clients, and the pattern is consistent: the teams that struggle are the ones who tried to make GA4 behave like UA. The teams that win treat GA4 as a new system with its own logic and set it up on its own terms. This article is structured as a comparison so you understand exactly what carried over, what changed, and what to configure, and then it becomes a working Google Analytics 4 setup tutorial you can execute this week.

Here is the quick orientation before we go deep.

• Understand The Shift: GA4 replaced the session-and-pageview model with an event-and-parameter model, and this single change explains 80% of the confusion.
• Do Not Migrate Config: You cannot copy UA goals or filters into GA4; you rebuild them as key events and explorations.
• Focus On Ecommerce Events: For a store, the purchase, add_to_cart, and begin_checkout events are where the money-relevant data lives.
• Verify Before You Trust: A property that receives data is not the same as a property that receives correct data.

GA4 vs Universal Analytics At A Glance

Before we take each side apart, here is the comparison table we walk clients through in the first scoping call. It maps the criteria that actually affect ecommerce decision-making, not the feature list Google publishes.

Criteria	Universal Analytics (retired)	Google Analytics 4 (current)
Data model	Sessions, pageviews, goals	Events with parameters, key events
Availability in 2026	Retired, no data processing	Live and the default property type
Ecommerce tracking	Enhanced Ecommerce plugin	Native ecommerce events (purchase, add_to_cart)
Cross-device tracking	Limited, cookie-based	Signals, user-ID, and modeling across web and app
Data retention (free tier)	Up to 50 months	2 to 14 months for exploration data
Reporting flexibility	Fixed standard reports	Explorations plus customizable reports
Privacy and consent	Weaker native controls	Consent Mode, data-deletion, IP anonymization by default
Cost of BigQuery export	Paid (360 only)	Free on standard tier

The single row that surprises people most is BigQuery export. In Universal Analytics, raw event export to BigQuery was a Google Analytics 360 feature costing six figures a year. In GA4 it is free on the standard tier, which quietly turned every small store into a company that can own its raw analytics data. When we scope this for clients, we treat that free export as a strategic asset, not a nice-to-have.

Section checklist:

• Confirm The Model: Know that GA4 is events-first before you open a single report.
• Note What Is Free Now: BigQuery raw export is available at no cost on standard GA4.
• Reset Expectations On Retention: GA4 exploration data expires far sooner than UA did, so export matters.
• Stop Reconciling To UA: Do not expect GA4 session or conversion numbers to match old UA reports.

The Case For Universal Analytics (And Why We Still Reference It)

Universal Analytics is gone, but its design still teaches useful lessons, and plenty of your team members will reason in its vocabulary for years. Understanding its strengths tells you what to consciously rebuild in GA4 so you do not lose capability during a migration.

Universal Analytics was, for most of a decade, the most widely understood analytics tool on the planet. Its session-based model matched how marketers instinctively think about a visit: someone lands, does some things, and leaves. Goals were trivial to configure. The Enhanced Ecommerce plugin, once installed, produced clean product-performance and shopping-behavior reports that a non-technical merchant could read without training.

Advantages:

• Intuitive Session Model: The visit-based structure matched how most marketers naturally describe user behavior, lowering the learning curve to near zero.
• Mature Ecosystem: Nearly every plugin, theme, and tag manager template supported UA, so implementation was often a five-minute job.
• Long Data Retention: Standard properties kept user and event data for up to 50 months, making year-over-year analysis straightforward.
• Stable Standard Reports: The fixed report layouts meant two people looking at the same data saw the same thing.

Limitations:

• Fully Retired: There is no path to use UA in 2026; it processes no data and the interface is gone.
• Weak Cross-Device Truth: Its cookie-based approach struggled with the multi-device, privacy-restricted reality of modern shopping.
• Bolt-On Ecommerce: Enhanced Ecommerce was a plugin layered on top, not a native part of the data model, so gaps were common.

The takeaway is not nostalgia. It is a checklist of capabilities to deliberately reproduce in GA4 so a migration is a net upgrade rather than a downgrade in disguise.

• Reproduce Goals As Key Events: Every UA goal maps to a GA4 key event you must define manually.
• Rebuild Ecommerce Reporting: Recreate product-performance and shopping-behavior views inside GA4 explorations.
• Preserve Historical Data: Export UA history to a spreadsheet or BigQuery before it is unrecoverable.
• Retrain The Team: Budget a few hours to move people off session-thinking and into event-thinking.

The Case For Google Analytics 4

GA4 is not simply a newer UA. It is a rearchitected product built for a world of multiple devices, tightening privacy law, and machine learning. Once you accept the event model, the platform is genuinely more powerful for ecommerce than UA ever was.

Everything in GA4 is an event. A pageview is an event. A purchase is an event. Scrolling 90% down a product page is an event if you want it to be. Each event carries parameters, small pieces of context like the item name, the value, or the currency. This is why the same setup that tracks your store also tracks your mobile app with almost no additional modeling: the data model does not care where the event came from.

Advantages:

• Unified Web And App Data: A single GA4 property can measure a website and a native app together, giving one honest view of a customer who shops across both.
• Free Raw Data Export: BigQuery export at no cost lets you own event-level data and build custom attribution or blend it with your CRM.
• Privacy-Ready By Design: Consent Mode, default IP anonymization, and data-deletion controls make GDPR and similar compliance far more manageable.
• Predictive And Flexible: Built-in predictive metrics like purchase probability, plus fully custom explorations, replace the rigid UA report menu.

Limitations:

• Steeper Learning Curve: The interface and vocabulary confuse anyone expecting UA, and misreading reports is common in the first month.
• Manual Ecommerce Setup: Native ecommerce events are powerful but must be implemented correctly; a broken purchase event silently poisons revenue reporting.
• Short Default Retention: Free-tier exploration data expires in as little as two months unless you export it, which catches teams off guard.

The honest summary is that GA4 rewards deliberate setup and punishes lazy setup harder than UA ever did. That is precisely why a proper Google Analytics 4 setup tutorial is the highest-leverage hour you will spend on analytics this quarter.

• Embrace Events: Stop looking for pageviews and start defining the events that matter to your revenue.
• Turn On BigQuery Early: Enable free export from day one so you never lose granular history.
• Configure Consent Mode: Set up consent signals before you drive paid traffic to protect data quality and compliance.
• Extend Retention: Bump exploration data retention to the 14-month maximum in admin settings immediately.

The Google Analytics 4 Setup Tutorial: Step By Step

Here is the working part. This is the sequence we run for ecommerce clients, ordered so that each step builds on the last. Treat it as a Google Analytics 4 setup tutorial you execute top to bottom rather than a menu you pick from.

Step	What You Do	Effort	Timeframe	Expected Outcome
1	Create the GA4 property and data stream	Low	10 minutes	A property ID and a web stream measurement ID
2	Install the tag (GTM or gtag.js)	Low to medium	20 to 40 minutes	Live pageview data in real-time reports
3	Enable enhanced measurement	Low	5 minutes	Automatic scroll, outbound click, and search events
4	Implement ecommerce events	Medium to high	2 to 8 hours	purchase, add_to_cart, begin_checkout firing correctly
5	Mark key events	Low	15 minutes	Conversion-worthy events flagged for reporting
6	Link ads and Search Console	Low	20 minutes	Cross-platform attribution and organic query data
7	Enable BigQuery export and extend retention	Low	15 minutes	Raw event ownership and 14-month retention
8	Verify with DebugView and test purchases	Medium	1 to 2 hours	Confidence that revenue numbers are accurate

Step one is creating the property. In the GA4 admin, create a new property, set your reporting time zone and currency correctly the first time because changing currency later does not retroactively fix historical revenue, then add a web data stream for your store domain. Copy the measurement ID that starts with G-.

Step two is installing the tag. You have two clean paths. The direct gtag.js snippet goes in your site head and is the fastest for a single simple store. Google Tag Manager is the option we recommend for any store that will grow, because it separates tag logic from code and lets a marketer manage tracking without a developer deploy. On Shopify specifically, the native Google and YouTube channel or a GTM install both work; pick one method and do not run two, or you will double-count.

Step three is enhanced measurement, which is a single toggle in the data stream settings that automatically captures scrolls, outbound clicks, site search, video engagement, and file downloads. It is free intelligence, so leave it on unless you have a specific reason not to.

Step four is the one that separates real ecommerce analytics from a vanity dashboard. GA4 has a defined set of recommended ecommerce events, and the four that matter most to revenue are view_item, add_to_cart, begin_checkout, and purchase. Each must be sent with the correct parameters: the items array, the value, the currency, and for purchase, a unique transaction_id so refunds and duplicates can be reconciled. On a platform like Shopify, much of this can be wired through the native integration or a data layer, but we still verify every event by hand because the difference between a purchase event that fires and one that fires with the right value is the difference between trustworthy revenue and fiction.

Step five is marking key events. In GA4, an event only counts as a conversion when you flag it as a key event. At minimum, mark purchase. Depending on your funnel, you may also mark begin_checkout or a newsletter signup, but resist flagging too many, because ten key events means none of them mean anything.

Steps six through eight are the multipliers. Link Google Ads so conversions flow back for bidding, link Search Console for organic query context, enable the free BigQuery export, extend data retention to 14 months, and then verify everything in DebugView by running a real test transaction end to end.

Section checklist:

• Set Currency First: Fix reporting currency at property creation because it does not correct historically.
• Choose One Tag Method: Run either gtag.js or GTM or the native channel, never two at once.
• Verify The Purchase Event: Confirm value, currency, and transaction_id on a real test order.
• Mark Only True Conversions: Flag a small, meaningful set of key events, led by purchase.
• Link Ads And Search Console: Connect both so attribution and organic data are complete.
• Turn On Free Export: Enable BigQuery and extend retention to 14 months on day one.

In GA4, a property that receives data is not the same as a property that receives correct data, and telling the two apart is the entire job.

Our Ecommerce GA4 Setup Framework: The CLEAR Method

Over dozens of implementations we developed a repeatable framework so nothing falls through the cracks. We call it CLEAR, and it is the sequence we run whether we are building a new store or auditing an inherited mess.

Configure the foundation. Property, stream, currency, time zone, and one clean tag install. Get this wrong and everything downstream inherits the error.

Layer the events. Enhanced measurement first for the free wins, then the four ecommerce events with correct parameters, then any custom events your funnel needs.

Enable the connections. Google Ads, Search Console, and BigQuery export, so GA4 stops being an island and starts feeding the rest of your stack.

Assign the key events. Deliberately choose the small set of events that represent real business outcomes, and mark nothing else.

Reconcile and verify. Test a live purchase, compare GA4 revenue to your platform’s own order data, and confirm they land within an acceptable margin before you trust a single report.

We ran this exact framework for Metalne Police, a Croatian retailer selling metal shelving and storage where every product is bought on numbers: dimensions, load capacity per shelf, total system weight. We built them a brand-new Shopify store and treated specifications as the interface, letting a first-time visitor go from “I need shelves for my garage” to a completed order without ever calling anyone. Clean structured product data feeds clean ecommerce events, and the analytics setup could actually attribute which space-and-use-case categories drove revenue. The store increased conversions by 18% compared to the previous setup, and the CLEAR sequence is part of why the reporting behind that number is trustworthy rather than guesswork.

As their CEO Davor Katalinic put it: “Our clients are usually old school and they need a very straightforward and easy way of online shopping. Presta delivered exactly that!”

Framework checklist:

• Configure: Lock the property foundation before touching anything else.
• Layer: Add enhanced measurement, then the four ecommerce events, then custom events.
• Enable: Connect Ads, Search Console, and BigQuery to break the data silo.
• Assign: Mark only the key events that map to real revenue outcomes.
• Reconcile: Verify GA4 revenue against platform order data before trusting reports.

Want Your Analytics Built Right The First Time?

If reading a Google Analytics 4 setup tutorial has made you realize your current tracking is either missing or quietly broken, this is exactly the problem our Startup Studio solves in days, not months. Presta’s Startup Studio builds and instruments ecommerce stores end to end, from a clean Shopify build to verified GA4 ecommerce events, BigQuery export, and dashboards your team can actually read, so you stop guessing which channels make money. If you want analytics you can trust from launch day, tell us what you are building and we will scope it with you.

Which Should You Choose

The literal choice between GA4 and Universal Analytics does not exist anymore, so the real decision is how deeply to invest in your GA4 setup. Here is the framework we map to specific situations.

What if you are launching a brand new store?

Build GA4 correctly from day one and enable BigQuery export immediately. New stores have the advantage of zero legacy mess, so you can implement the four ecommerce events cleanly, mark key events deliberately, and start accumulating raw data you own from your first order. This is the cheapest possible time to do it right, because retrofitting later costs 3 to 5 times the effort. Since a new store almost always means Shopify or a similar platform, this is also where a proper build and instrumentation pay off fastest, and our journey of a new website walks through how we sequence build and measurement together.

What if you inherited a half-configured GA4 property?

Audit before you touch anything. Roughly 40% of the properties we open are missing ecommerce events or have a broken purchase event, and the danger is that the data looks plausible while being wrong. Run a structured audit: verify the tag fires once, confirm the purchase event carries value and transaction_id, check whether key events are marked sensibly, and reconcile a week of GA4 revenue against actual orders. Our systematic approach to debugging is the same discipline we apply to analytics audits, because a broken data pipeline is just a bug you cannot see.

What if analytics is not your team’s strength?

Then the honest recommendation is to get help for the setup and keep the day-to-day yourself. GA4 configuration is a one-time high-skill task; reading the resulting reports is a low-skill recurring task. Pay for expertise where it compounds, which is the setup and verification, and own the ongoing monitoring. This is exactly the logic behind hiring an experienced agency for the build and keeping operations in-house.

Decision checklist:

• New Store: Instrument GA4 correctly from order one and enable export immediately.
• Inherited Property: Audit and reconcile before you trust any existing report.
• Weak Analytics Skills: Buy the setup, own the monitoring.
• Multi-Device Business: GA4 is a clear upgrade over anything cookie-based UA offered.
• Privacy-Sensitive Market: Configure Consent Mode before driving paid traffic.

Measuring Success: 30, 60, And 90 Day KPIs

A GA4 setup is only worth doing if you can prove it is working. Here is the timeline of outcomes we hold ourselves and our clients to, so you know whether your implementation is on track or quietly failing.

Timeframe	Milestone	KPI To Verify	Healthy Target
Day 30	Data flowing and events firing	GA4 revenue vs platform revenue	Within 5% variance
Day 30	Key events marked	Purchase counted as conversion	100% of true orders captured
Day 60	Attribution connected	Ads and Search Console linked	Both feeding data, no gaps
Day 60	Funnel visible	begin_checkout to purchase rate	Baseline established and stable
Day 90	Insight-driven decisions	Reports used in a real decision	At least one channel budget reallocated
Day 90	Data ownership	BigQuery export accumulating	90 days of raw events stored

The first 30 days are about trust. If GA4 revenue and your platform’s own order total do not agree within roughly 5%, stop everything and fix the purchase event before you look at anything else. In our experience, this single reconciliation catches 90% of setup errors.

Days 30 to 60 are about connection. A GA4 property that is not linked to Google Ads and Search Console is a diary, not a dashboard. Once those links are live, you can start seeing which campaigns and which organic queries actually drive the events you marked as key events.

Days 60 to 90 are about decisions. The whole point of analytics is to change what you do. By day 90 you should have made at least one real decision from the data, reallocated a budget, killed a channel, or doubled down on a winning product category. If your GA4 property has never changed a decision, it is a cost, not an asset.

KPI checklist:

• Reconcile Revenue: Confirm GA4 and platform revenue agree within 5% by day 30.
• Confirm Conversions: Verify every true order counts as a purchase key event.
• Connect Attribution: Link Ads and Search Console by day 60.
• Baseline The Funnel: Establish a begin_checkout to purchase rate to improve against.
• Prove Impact: Point to at least one decision the data drove by day 90.

Common GA4 Setup Mistakes We Fix Repeatedly

Because we audit so many properties, the failure modes are predictable. Here are the ones that cost the most money, ranked by how often we see them.

Mistake	Why It Hurts	How Often We See It
Double-firing tags	Doubles sessions and revenue, making everything look inflated	Very common
Purchase event missing value	Revenue reports read zero or wildly wrong	Common
Too many key events	Dilutes conversion signal for ad bidding	Common
No BigQuery export	Loses granular data forever after retention window	Very common
Currency set wrong at creation	Corrupts all historical revenue, uncorrectable	Occasional
Consent Mode not configured	Data loss and compliance exposure in EU markets	Common

The double-firing tag deserves special attention because it is the most likely to fool you. When a store installs GA4 through both a native platform channel and Google Tag Manager, every purchase fires twice, and suddenly your revenue looks 100% higher than reality. We have watched founders make a hiring decision on doubled numbers. Always confirm each event fires exactly once in DebugView.

Section checklist:

• Check For Doubles: Confirm each event fires exactly once in DebugView.
• Validate Purchase Value: Ensure the purchase event always carries a real value and currency.
• Prune Key Events: Keep the key event list short and meaningful.
• Enable Export Now: Turn on BigQuery before the retention window eats your history.
• Lock Currency Early: Set the correct reporting currency at property creation.

Where GA4 Is Heading: Agentic Commerce And AI-SEO

Setting up GA4 correctly is not just about today’s reports. The measurement layer you build now is the foundation for the next shift in ecommerce: AI-driven and agentic commerce, where autonomous agents and AI assistants increasingly mediate discovery and purchase. Clean, structured event data is what makes your store legible to those systems, and it is what lets you measure a buying journey that no longer looks like a tidy session.

We write about this shift in depth in our guide to agentic commerce and Google’s Universal Commerce Protocol, and it connects directly to analytics: if you cannot cleanly attribute a purchase today, you certainly will not attribute an agent-mediated one tomorrow. The same discipline that makes your GA4 purchase event trustworthy is the discipline that will make your store measurable in an agentic commerce integration, and it overlaps heavily with how a Gemini and Shopify integration surfaces your catalogue to AI shopping assistants.

The practical point: the event model that makes GA4 feel harder than UA is exactly the model that makes it future-proof. Events with rich parameters are the lingua franca of modern and AI-mediated commerce, so setting them up well is an investment that keeps paying.

Section checklist:

• Structure Product Data: Clean specs and parameters feed both GA4 and AI shopping surfaces.
• Own Your Raw Events: BigQuery export is your bridge to future attribution models.
• Think Beyond Sessions: Design measurement for journeys that span devices and agents.
• Keep Events Rich: Populate item, value, and currency parameters fully, not minimally.

The Verdict Table

Here is how the two options score across the criteria that matter, given the 2026 reality.

Criteria	Winner
Availability today	GA4 (UA is retired)
Ecommerce data model	GA4 (native events)
Cross-device accuracy	GA4
Free raw data export	GA4 (BigQuery included)
Privacy and compliance	GA4
Ease for a non-technical reader	Universal Analytics (but irrelevant now)
Historical data retention	Universal Analytics (up to 50 months)
Future-readiness for AI commerce	GA4

The verdict is not close and it is not really a choice. Universal Analytics won on approachability and raw retention, but it no longer exists as a live product, and even if it did, its cookie-based, session-bound model is fundamentally mismatched to how people shop in 2026. GA4 wins every category that matters going forward. The only real decision left is whether you set it up carefully or carelessly, and that decision determines whether your analytics is an asset you steer by or a dashboard you quietly distrust.

If you are just getting started, prioritize the foundation and the four ecommerce events above everything else: get the property, currency, tag, and a verified purchase event right, and enable BigQuery on day one so you never lose history. Everything else can be added later, but a broken purchase event poisons every report from the start. If you are auditing something that already exists, do not touch a single setting until you have reconciled a week of GA4 revenue against your actual orders, because the most dangerous GA4 property is not the empty one, it is the confidently wrong one.

Next Steps:

• Run A Reconciliation: Compare last week’s GA4 revenue against your platform’s order total and note the variance.
• Verify Your Purchase Event: Place one test order and confirm value, currency, and transaction_id in DebugView.
• Enable BigQuery Export: Turn it on today so raw event history starts accumulating immediately.

Frequently Asked Questions

Where can I find a GA4 setup tutorial?

The most authoritative Google Analytics 4 setup tutorial is Google’s own Analytics Help documentation, which is kept current and covers the property creation, data stream, and enhanced measurement steps in detail. It is the reference we point clients to for the platform-neutral basics, and it is free.

That said, general tutorials stop short of the ecommerce-specific work that actually drives revenue reporting. The generic guide will show you how to create a property and install a tag, but it will not walk you through wiring the four ecommerce events with correct parameters on your specific platform, which is where most setups break. For that, look for platform-specific guides, Shopify’s own documentation if you are on Shopify, and verify everything against Google’s ecommerce event reference.

The pattern we recommend is: use Google’s documentation for the platform-neutral setup, use your ecommerce platform’s docs for the integration, and always verify against the recommended events reference so your parameters match what GA4 expects. That combination gets you a setup that reports accurate revenue rather than a property that merely receives traffic.

What is the easiest way to set up GA4?

The genuinely easiest path depends on your platform. If you are on Shopify, the native Google and YouTube channel handles the base install and much of the ecommerce event wiring for you, which for a simple store can be the fastest route to live data. For most other cases, Google Tag Manager is the easiest scalable option because it lets you manage tags without editing code.

Easy, though, is not the same as complete. The base install is the simple 20-minute part; the part that determines whether your data is trustworthy is verifying that ecommerce events fire correctly with the right value and currency. So the honest answer is that setting up GA4 to receive some data is easy, and setting it up to report accurate revenue takes a bit more care.

If you want the least painful path that still ends in trustworthy data, follow a step-by-step sequence like the CLEAR framework in this article: configure the foundation, layer the events, enable connections, assign key events, and reconcile. Skipping the reconcile step is what makes an easy setup a wrong one.

Are there step-by-step GA4 setup tutorials?

Yes, and this article is one of them; the step-by-step section above takes you from property creation through verification in eight ordered steps with expected outcomes and timeframes. Following a numbered sequence matters more in GA4 than it did in UA because the steps genuinely build on each other, and skipping ahead tends to produce a property that looks configured but reports wrong numbers.

The steps that people most often skip in a step-by-step tutorial are the least glamorous ones: enabling BigQuery export, extending data retention, and running a real test purchase to verify. Those three steps are exactly the ones that separate a professional setup from a hobbyist one, so do not treat them as optional.

If you prefer a framework you can memorize rather than a long list, use CLEAR: Configure, Layer, Enable, Assign, Reconcile. It compresses the full step-by-step tutorial into five phases you can hold in your head while you work.

How is GA4 different from Universal Analytics for ecommerce?

The core difference is the data model. Universal Analytics used sessions and layered Enhanced Ecommerce on top as a plugin, while GA4 treats ecommerce events like purchase and add_to_cart as first-class native events with parameters. This means GA4 ecommerce tracking is more integrated but requires deliberate implementation of each event.

Practically, that changes how you read reports. UA gave you fixed shopping-behavior and product-performance reports out of the box; GA4 gives you the raw events and expects you to build or customize the views, often through explorations. It is more flexible and more work, and the flexibility only pays off if the underlying events are configured correctly.

The other big ecommerce difference is data ownership. GA4’s free BigQuery export lets you keep every raw ecommerce event, which UA reserved for its expensive 360 tier. For a growing store, that free export is arguably the single most valuable difference, because it future-proofs your attribution.

Will my GA4 numbers match my old Universal Analytics reports?

No, and expecting them to is one of the most common sources of frustration we see. GA4 counts sessions, users, and conversions differently from UA, so even a perfect setup will show numbers that do not reconcile with old UA reports. This is expected behavior, not a bug in your setup.

The differences come from real methodological changes: GA4 defines sessions differently, models some data where consent is missing, and counts conversions per event rather than once per session by default. Trying to force the two to match wastes time and usually leads people to distrust a correct GA4 setup.

The right mindset is to treat GA4 as a fresh baseline. Compare GA4 to GA4 over time, compare GA4 revenue to your ecommerce platform’s actual orders for accuracy, and stop comparing GA4 to a retired tool. Once you make that shift, the numbers become useful again.

When does it make sense to bring in Presta’s Startup Studio for GA4 setup?

Candidly, not every store needs an agency for this. If you are running a simple Shopify store, you are comfortable following a step-by-step tutorial, and you have an hour or two to verify events in DebugView, you can absolutely do a correct GA4 setup yourself, and you should. The platform is accessible enough that a careful operator gets a trustworthy result.

The threshold where it becomes worth bringing us in is when the cost of wrong data starts to exceed the cost of expert setup. That happens in a few specific situations: when you are spending meaningfully on paid ads and bidding on GA4 conversions, so a broken event directly wastes budget; when you are running across web and app or multiple markets and currencies, where the setup complexity multiplies; when you have inherited a property you suspect is wrong and cannot tell; or when you are building a new store and want measurement instrumented correctly from launch rather than retrofitted.

In those cases, the math usually favors expert setup, because a one-time configuration by people who have done it dozens of times prevents months of decisions made on bad data. Our Startup Studio typically instruments a store’s full GA4 and BigQuery pipeline in days, and the payback is every ad dollar you stop wasting on channels you were misattributing. If any of those thresholds describe you, that is the moment to reach out rather than to keep second-guessing your dashboard.

How long does a proper GA4 ecommerce setup take?

For a straightforward single-store, single-currency Shopify site with a clean theme, a complete and verified setup is realistically a half day to a full day of focused work, including testing. The base property and tag are quick; the ecommerce events and verification are where the time goes.

For a more complex situation, a custom-coded store, multiple currencies, a web and app combination, or a migration from a messy inherited property, plan for two to five days. The extra time is almost entirely in event implementation and reconciliation, not in the platform basics.

The one part you should never compress is verification. Whatever the total timeline, budget at least one to two hours for placing a test order and confirming that revenue reconciles, because that step is what converts a setup that exists into a setup you can trust.

Sources

• Google Analytics Help: Universal Analytics sunset
• Google Analytics Help: GA4 recommended ecommerce events
• Google Analytics Help: Set up a GA4 property
• Google Analytics Help: BigQuery export for GA4
• Google Analytics Help: Enhanced measurement
• Google Analytics Help: Key events and conversions