GA4 eCommerce Hacks: Top Features You Should Be Using

No. Enhanced Ecommerce was a Universal Analytics feature with its own toggle, and GA4 has no equivalent setting. GA4 ecommerce reports fill in only when your site sends the reserved events with a properly structured items array, principally view_item, add_to_cart, begin_checkout and purchase. There is no admin screen that turns this on. If your ecommerce reports are empty, the gap is in your tagging or platform integration, and hunting through GA4 settings will not find it. Run a test purchase with DebugView open and you will know within minutes which of the four core events your store actually sends and which are missing.

The first two are right and the third is wrong. GA4's reserved ecommerce event is purchase, not purchase_complete. Send purchase_complete and GA4 accepts it as an ordinary custom event that never appears in ecommerce reporting, which is exactly how stores end up with functioning tags and empty revenue reports. Reserved names are not stylistic preferences, they are what the standard reports key on. Reserve custom event names for genuinely custom actions such as a size guide opening or a video completing.

Actions specific to your business that no reserved event covers and that you would change something because of. A size guide opening, a stock notification request, a finance calculator being used, a specification sheet downloaded, a live chat opened from a product page. Each of those tells you something about hesitation or intent that the standard funnel misses. What is not worth tracking is every click on the site, which fills your custom dimension quota and produces reports nobody opens twice.

Standard GA4 properties allow 50 event scoped custom dimensions, 25 user scoped and 50 custom metrics, with higher ceilings on Analytics 360. Teams hit those limits faster than expected because it is tempting to register every parameter being sent. Register only what you will genuinely segment or report on. Parameters are collected whether or not you register them and remain readable in the BigQuery export, so the quota is about reporting convenience rather than data capture. Plan the list before you start filling it.

Good as an alarm, weak as an explanation. The system reliably notices that a metric moved unusually, which genuinely helps when a tag breaks or a campaign misfires. It knows nothing about your promotions, stock outages, price changes or seasonality, so its suggested causes are often wrong. Treat an insight as a prompt to look, never as a finding to report. Pair it with a custom alert on purchase event volume, which catches the failure mode that costs the most and is otherwise invisible for weeks.

Check before you plan around them. GA4 requires at least 1,000 returning users who triggered the relevant condition and 1,000 who did not, within a seven day window over the previous 28 days, with model quality sustained over time, and only purchase and in-app purchase events are supported. Plenty of healthy mid sized stores never qualify. Verify in the predictive section of the Audience Builder. Where you do not qualify, behavioural audiences built from observed intent work at any volume and are easier to justify to a client.

The BigQuery export, comfortably. It is free to link, costs only storage and queries, and is the only way to keep event level data past the 14 month retention ceiling on standard properties or to join GA4 behaviour against your own order, margin and returns data. The catch is that it captures nothing retrospectively, so enabling it early is nearly free and enabling it late costs history permanently. Most properties we audit have never turned it on. Enabling it takes a few minutes in the admin panel and needs no engineering work, which makes the usual reason for skipping it hard to defend.

Because that is GA4's default and it is rarely changed. Standard properties can go to 14 months, with 26, 38 and 50 month options only on Analytics 360. The limit governs the user and event level data behind Explorations, funnels, path analysis and custom segments, while aggregated standard reports are kept longer. The setting is not retroactive, so raising it today preserves the next 14 months and recovers nothing from before. It is the single quickest configuration win available.

Partially. When you create an audience GA4 can backfill from up to 30 days of prior data where the conditions permit, but membership mostly accrues from creation onward. That makes audience building a task to do early rather than the week a campaign launches. Define the segments you know you will want, such as cart abandoners, repeat purchasers and high value browsers, and let them accumulate. An audience created the day before a campaign is nearly empty when you need it.

Not directly, and expecting it to is a common gap. Core Web Vitals field data lives in Search Console and PageSpeed Insights, measured against thresholds of 2.5 seconds for Largest Contentful Paint, 200 milliseconds for Interaction to Next Paint and 0.1 for Cumulative Layout Shift. What GA4 shows is the consequence in engagement and conversion. Google and Deloitte measured a 0.1 second mobile speed improvement lifting retail conversion 8.4 percent and average order value 9.2 percent, which is what you should expect to see move.

Duplicate purchase events, usually from running a platform integration and a Tag Manager tag simultaneously, or from a confirmation page that refires on refresh. Revenue and conversion rate inflate, and every optimisation decision afterwards is measured against a distorted baseline. Send transaction_id with every purchase event so GA4 can deduplicate, then audit which system is actually sending it and disable the other. Do this before you trust a single revenue figure, because no downstream analysis survives it. A quick check is to compare GA4 transactions against your platform's order count for a single day, since duplication shows up immediately at roughly double.

Fix event names and verify a real test purchase in DebugView, since everything else depends on that. Raise data retention to 14 months and link BigQuery, because neither is retroactive and both cost history every day they are delayed. Build one segmented funnel exploration and set an alert on purchase volume. Then add custom events for the two or three hesitation moments specific to your store. Audiences and predictive features are worth having, and worth nothing on top of unreliable data.