GA4 Guide: Setup, Events, and Reports for Online Stores

GA4 (Google Analytics 4) helps you understand how people find your website, what they do there, and which visits lead to valuable actions. For an online store, that means connecting traffic with product views, checkout activity, and measured purchases.

Getting a dashboard to show visitors is the easy part. Knowing whether it captured the right purchases, credited the right traffic source, and used the right revenue definition takes more care.

This Google Analytics 4 tutorial walks through that work: a useful setup, events worth collecting, reports worth reading, and checks to run before changing your marketing or store. The examples use US dollars, but the measurement principles apply to stores in other markets too.

What is GA4?

Google Analytics 4 is Google’s event-based analytics platform for measuring activity on websites and apps. An event records an action, such as viewing a page, adding an item to a cart, or completing a purchase. Parameters provide the details, such as the page URL, product ID, or purchase value.

The basic structure has four parts:

• Account: the container for your business’s Analytics properties and access.

• Property: the reporting space with its own settings and collected data.

• Data stream: the source sending website or app activity into that property.

• Event: an action recorded from that source, with relevant details attached.

A web data stream and an installed tag are both needed to start website collection. Creating a property alone doesn’t measure your store.

Universal Analytics is retired. Its historical data cannot be imported into GA4 as previously collected activity, so an old migration tutorial won’t establish a working setup today. Google’s migration reference explains that historical-data boundary.

What are the practical benefits of Google Analytics 4?

Google Analytics 4 benefits store owners when it turns a business question into something measurable. Start with the question, then decide which events and report can answer it.

• Compare acquisition channels. Check whether search, email, or paid campaigns bring visits that lead to the actions you care about.

• Inspect shopping behavior. Measure where tracked users move from product interest to checkout and purchase.

• Evaluate landing pages. Look beyond visits to the engagement and outcomes associated with each entry page.

• Build useful comparisons. Compare devices, campaigns, or audiences to narrow an investigation instead of reading one overall average.

Those capabilities require a working implementation. Analytics won’t improve a checkout by itself, and a traffic increase doesn’t establish profitable growth.

How does GA4 event tracking work?

Event tracking combines activity collected by the base setup with actions your site or integration must send explicitly. Knowing the difference prevents a common mistake: assuming that working page views mean working ecommerce tracking.

Which events are automatic, and which need setup?

The base tag collects some activity automatically. Enhanced measurement adds supported interactions such as scrolls, outbound clicks, site search, and certain video, download, or form actions when configured.

Recommended ecommerce events use Google’s defined names and parameters. Your store integration or custom implementation must send them; turning on enhanced measurement does not create a complete sales setup.

Use Google’s recommended event definitions when mapping these actions. A custom event such as order_done doesn’t become a standard purchase simply because its name sounds similar.

What is a key event?

A key event is a collected event marked as important to the business. A completed purchase is a useful outcome; a page view or scroll usually answers a different question.

Google distinguishes key events from conversions. Key events support behavioral reporting, while conversions support advertising measurement and bidding. Review the linked Ads configuration separately rather than assuming a key-event label sets up every campaign goal.

What does engagement rate tell you?

Engagement rate is the share of sessions classified as engaged. At the default threshold, an engaged session lasts longer than 10 seconds, contains a key event, or has at least two page or screen views. Bounce rate is the inverse, according to Google’s metric definitions.

Engagement measures activity, not buyer intent. A higher rate isn’t automatically better if you changed which events count as important.

How do you set up GA4 for an online store?

A useful setup needs a measurement goal, the right property and stream, a collection route, and ecommerce events. Before starting, identify who can edit the website integration, Analytics property, and Tag Manager container if one is used.

Step 1: decide what you need to measure

Write three questions your business needs answered. For example: which campaigns bring purchases, which products attract cart activity, and where mobile shoppers leave checkout.

Map each question to its required events. Agree on what counts as a completed order, how refunds are recorded, and which system owns the order record. That gives the person implementing tracking a clear specification.

Step 2: create the property and web data stream

Open Google Analytics and create an account if your business needs one. In Admin, create a property with the correct reporting time zone and currency, then select the relevant business details and objectives.

Add a Web data stream for the website and record its measurement ID. Check these settings against your store reporting because mismatched day boundaries and currencies complicate comparisons.

Follow Google’s website setup instructions for the current screens and access requirements. Default report collections can differ according to the business objectives selected.

Step 3: choose one primary collection route

Use the store platform’s supported integration, Google Tag Manager, or a direct Google tag installation. Check what is already installed before adding another route; two implementations can send the same activity twice.

For Shopify or WooCommerce, verify the events your chosen integration actually supplies. Our Shopify GA4 guide and WooCommerce Google Analytics guide provide platform-specific context.

With Tag Manager, Google’s current setup guide documents a Google Analytics destination route and an alternate Google Tag route. In the newer route, use Tags > New > Tag Configuration > Google Analytics > Google Analytics, then enter the relevant tag ID.

The alternate route uses Google Tag as the tag type. Follow the route available in your container, preview it with Tag Assistant, and publish the tested version; saving a tag isn’t the same as publishing it.

Step 4: add ecommerce measurement

Check that the integration sends the shopping events you need, with consistent product identifiers and valid order details. If it doesn’t, have the implementer add the missing measurement rather than adding a second full integration without a plan.

Google’s Ecommerce purchases documentation makes the distinction explicit: ecommerce activity must be sent before the corresponding sales reports can show it. Verify the actual output of a plugin or platform connection instead of relying on its installed status.

Review the property’s key-event settings so the intended purchase outcome is included. Keep exploratory actions, such as reading a page, separate from the outcome you want to optimize.

What should a purchase event include?

A purchase event needs order-level and item-level details that describe the same completed order. Start with a unique transaction ID, a numeric value, the currency, and an items array. An event appearing in a list doesn’t prove that its revenue parameters are valid.

How do order value, tax, and shipping fit together?

Consider a hypothetical order with two items priced at $40 each after discounts, $6 shipping, and $6 tax. The customer pays $92, but the item-value calculation is $80.

These are illustrative parameters for that purchase event:

{
  "transaction_id": "EXAMPLE-ORDER-1001",
  "currency": "USD",
  "value": 80,
  "shipping": 6,
  "tax": 6,
  "items": [
    {
      "item_id": "TEE-NAVY-M",
      "item_name": "Navy tee",
      "price": 40,
      "quantity": 2
    }
  ]
}

The implementation must supply each real order’s values; this is a parameter example, not a complete installation. Google’s ecommerce collection guide separates item value, shipping, and tax.

For a discounted item, send the price actually charged. Google’s discount guidance explains that adding a discount parameter does not make Analytics subtract it from price. Check the arithmetic rather than applying the same discount twice.

Why do transaction IDs and refunds matter?

Use a different, nonempty transaction_id for each order and retain it if the same order is sent again. Google documents purchase deduplication for web streams; a reused or empty ID can cause purchases to be undercounted.

Send the relevant refund event when a measured order is refunded, using its original transaction ID and appropriate item details. Refunding an order in your store doesn’t prove your analytics integration sent the matching event.

How do you check that tracking works?

Validate a complete shopping sequence and its parameters before trusting the sales reports. Then allow processing time before comparing the published numbers.

Check the event sequence and payload

Use Tag Assistant for a Tag Manager implementation and enable debug mode for the device you’re testing. Open Admin > Data display > DebugView and inspect the arriving events.

Google’s DebugView instructions explain that consent and privacy controls can affect visibility. Confirm the device and collection conditions before treating an empty debug view as a broken installation.

For one controlled test flow, check:

• Product view, cart, checkout, and purchase actions send the intended event names.

• The same product identifier follows the item through those events.

• The purchase includes a unique order ID, numeric value, currency, and items.

• The implementation doesn’t send a new purchase each time the confirmation page is refreshed.

• Shipping, tax, discounts, and refunds follow the agreed measurement rules.

Keep the test order ID and expected values so you can check that specific record later. Use the platform’s test workflow where appropriate, and account for any test-data filters when reviewing results.

Separate debugging from processed reporting

Realtime and DebugView help confirm collection. They don’t certify the final revenue report.

Google says data processing can take 24 to 48 hours, and reports can change during that period. Compare completed periods with matching dates and definitions; don’t diagnose a sales collapse from an unfinished day’s data.

Which GA4 reports should you read first?

Choose a report according to the question you need answered. Opening every report usually creates more comparisons than decisions.

Report collections can vary between properties or be customized. Search for the report name if your menu differs, then check the date range, filters, and selected metrics.

Our GA4 ecommerce reports guide explains individual reports in more depth. Use it when you need a report walkthrough rather than another general introduction.

Why do acquisition reports show different answers?

User acquisition and Traffic acquisition have different scopes. One groups activity by how a user was first acquired; the other uses session-level acquisition information.

Suppose a shopper first finds your store through organic search and later returns through a tagged email campaign. If Analytics recognizes that shopper across both visits, the first-user source can remain organic search while the later visit has an email session source.

Neither view is inherently wrong. Select the one that matches your question, and don’t assume its revenue credit is identical to an advertising attribution report.

How can Explorations help you find a problem?

Explorations let you investigate a narrower question than a standard report answers. A funnel can examine a defined sequence; a path can show what happened before or after an event; a cohort can compare behavior among users grouped by a shared characteristic.

Build a funnel around one decision

Start with Explore > Funnel exploration and define the actions you want to examine. Google’s funnel documentation distinguishes closed funnels, which require entry at the first step, from open funnels, which allow entry later.

For a hypothetical closed funnel, assume 1,000 users view a product, 200 add to cart, 120 begin checkout, and 60 purchase in the defined sequence and period. That gives step progression of 20%, 60%, and 50%, respectively.

Half the included checkout users reached the purchase step. That identifies a place to investigate; it does not establish payment failure or prove a checkout redesign is needed.

Check event coverage first, then compare a useful breakdown such as mobile versus desktop. Look at the actual checkout only after confirming that the reported gap isn’t caused by missing events.

Keep the funnel definition consistent

Use the same entry rule, step sequence, date range, and timing conditions when comparing periods. Users, sessions, and event counts are different units; summing purchase events beside product-view users doesn’t create a valid funnel.

Write the question beside the exploration, such as “Did mobile checkout progression change after the release?” A saved chart without its definition is hard for another person to interpret.

Which settings and limits deserve attention?

Review checkout domains, retention, consent, and revenue definitions before treating the dashboard as a complete record of the business. These settings determine what is collected, how long some analysis remains available, and what comparisons mean.

Check checkout domains and payment referrals

If a journey crosses your own website and checkout domains, review Google’s cross-domain measurement instructions. The included pages need the same web-stream tag ID and a working linker configuration.

For a return from an external payment provider, review unwanted referrals. Preventing a payment-provider return from being classified as a referral is different from tracking activity inside that provider’s checkout. It won’t repair a missing purchase event.

Choose retention for the analysis you need

Standard properties offer two- or 14-month event-data retention settings. Google’s retention guidance distinguishes Explorations and funnel reports from ordinary aggregated reports, which aren’t affected in the same way.

Select a period that supports your intended analysis and data policy. Increasing retention later does not bring back data that has already been deleted.

Verify consent and ordinary event data

Consent mode communicates visitor choices to tags; it isn’t a consent banner. Check how your collection behaves when someone accepts or declines, including on the checkout path.

Modeled data is not the same as an observed order. Google’s behavioral modeling documentation sets eligibility requirements, so don’t assume modeling fills every gap in every property.

Keep customer names, email addresses, and order addresses out of ordinary event parameters, URLs, and campaign values. Follow Google’s PII guidance when reviewing those fields.

Compare matching revenue definitions

Before comparing Analytics with your store, align the period, time zone, currency, refunds, and treatment of tax and shipping. Check the exact metric, including whether it describes items, purchases, or broader revenue.

If purchases arrive without revenue, inspect the payload using Google’s missing revenue checklist. If only some orders are absent, compare identifiable records and collection conditions before assuming the difference is a fixed undercount.

Revenue also isn’t profit. Evaluate product costs, fees, shipping costs, and other business expenses before increasing spend on a channel that looks strong in a revenue report.

How do you connect website behavior with business results?

Use website analytics to investigate acquisition and on-site behavior, then use store and payment records to verify orders, refunds, and business outcomes. A customer can buy repeatedly or through another channel without every transaction being represented by the same tracked website journey.

When you sell through several stores or gateways, Putler can consolidate connected transaction data into revenue, customer, product, and subscription reporting. That adds a business view alongside your traffic analysis; it doesn’t fix missing site events or reproduce Google Ads attribution.

The ecommerce web analytics guide explains how to connect these questions. Keep the source records available when you need to reconcile transactions or bank deposits.

What should you do this week?

Start with one measurable business question and the events it requires. Verify a complete shopping flow, keep the order ID and expected values, and check the processed report after the collection window.

Then review one complete period, select the report with the right scope, and record one specific next action. If a release changes your tracking or checkout, repeat the verification before comparing performance.

Frequently asked questions