Your Google Analytics Is Recording Data. It's Just Not Telling You Anything.

Most stores we look at have Google Analytics installed, but almost none of them have it setup with a sales funnel.

It’s easy to think installed means “working” but adding GA4 just means you’re recording data. Analytics is more than recording sessions and pageviews.

In our post on the marketing metrics that actually matter, we set out a simple model: spend, visitors, leads, customers, revenue, and the cost to acquire a customer that falls out of those. The direction of your site lives in that top line view. Diagnosis lives one level down, in the reports that explain why a number moved. What we didn’t cover there is that for most ecommerce stores, the “diagnosis” layer isn’t something additional you need to buy. It’s very likely already sitting in your GA4 account, half-configured and never looked at.

The gap is almost always in the middle

Spend and Visitors are the easy half of the model. Ad platforms report spend, GA4 counts sessions, nobody has to do anything unusual to see either number.

Leads and Customers are where most setups fall over; GA4’s ecommerce tracking runs on a set of named events that have to fire as someone moves through your site. The four that matter most are view_item, add_to_cart, begin_checkout, and purchase. Most ecommerce platforms send some of these out of the box. Almost none of them send all four correctly. It’s entirely possible to have GA4 “installed” for years and still be missing the one event that would tell you where people are actually dropping off, especially if you’re running a custom build.

How to know? Open GA4, go to Admin, then Events, and look for those four names. If purchase isn’t there, GA4 has no visibility into revenue at all. Your store platform still knows what sold, but GA4 can’t tell you which channel or campaign actually drove it. If begin_checkout isn’t there, you have no visibility into cart abandonment at all, you only know sessions happened and some of them ended in a sale.

What the gaps between events actually tell you

Once those four events are genuinely flowing, something useful happens: the distance between any two of them stops being an abstract number and starts pointing at a specific part of the business.

If traffic looks healthy but add_to_cart is low relative to view_item, that’s not a traffic problem. It’s a product page problem: the offer, the images, the price, the description aren’t doing their job once someone’s actually looking at the item.

If add_to_cart looks fine but begin_checkout is where the drop happens, that’s cart friction. Shipping cost showing up as a surprise, a forced or high friction account creation step, a cart page that’s slow or confusing. People wanted the product. Something in the next step made them think it wasn’t worth it.

If begin_checkout is healthy but purchase drops away, the problem is the checkout itself. Payment method gaps, a broken step, missing trust signals at the exact moment someone’s about to hand over their card details.

Three drop-off points, three completely different conversations. None of that is visible from a single “conversion rate” number. It only becomes visible once the events between it are being tracked and you look at the shape of the drop-off, not just its total size.

A measurement plan is what makes this deliberate

The four core events give you the generic version of the funnel, but most stores have stages that don’t fit neatly into view_item, add_to_cart, begin_checkout, and purchase. A newsletter signup, a quote request, an account creation step, a size-guide interaction that predicts purchase intent, none of that is captured by default. That’s the actual job of a measurement plan: deciding which of those extra moments matter enough to track deliberately for your specific store, rather than accepting whatever a platform integration happened to ship with out of the box.

A caveat worth knowing once you start relying on this

Once this data is genuinely informing decisions, you need to understand what GA4 is actually counting and what it’s missing. Sampling on high-traffic properties and ad-blocker loss mean the numbers you see should only ever been read as a trend line, not a precise headcount. This is fine for the diagnostic work, a real gap between add_to_cart and begin_checkout will show up clearly whether the underlying numbers are off by a few percent or not. It becomes a problem if you start treating GA4 as a source of exact figures for anything financial. Treat it as directional tool, not a precision instrument, and you won’t get caught out.

Where it eventually connects to revenue, not just conversion

Everything above stops at the purchase event, which is a reasonable place to stop for a first pass. The next layer, once this is working, is connecting that first purchase to what happens afterwards: repeat orders, email-driven revenue, actual customer lifetime value by channel. That’s a different conversation, but it’s the same underlying principle.

If you’ve never actually checked whether your purchase event is firing correctly, do it before someone asks why conversions looked strange last month.