You came to investigate checkout. The data showed the problem is elsewhere. And then it showed that checkout has its own problem too – just not the one you thought.
What session_funnel measures – and what it doesn’t
session_funnel tracks one path: from the moment of landing on the site to placing an order. Five steps – session, product views, add to cart, checkout start, completion – and between every two steps a number that says what percentage of users took the next step.
Sounds simple. But in practice this funnel breaks the most common belief about conversion problems: that checkout is the culprit. Checkout is last in line, so attention naturally focuses on it. Yet the data often shows that the real leak starts much earlier.
Most stores think they have a checkout problem. session_funnel shows where the problem really is.
An important caveat: session_funnel measures behavior at the site level, not per product. If you want to know which specific products have an add-to-cart problem – there’s a separate tool for that (product_funnel). Here we look at the overall, per-channel picture.
Your funnel – where the money really leaks
Below is data from a real e-commerce store over 90 days. The result is representative at 206,960 sessions – statistically reliable.
206,960 sessions → 11,551 transactions · overall conversion 5.58%
Weakest stage: view_to_cart (20.09%). Of the 114k+ people who viewed a product, only 1 in 5 added it to the cart. This is where 4× more customers drop off than in checkout itself. Checkout_to_transaction is 62.66% – real, but not the biggest problem.
This is a classic result that surprises most store owners. Intuition says: “I have a checkout problem – too many people drop off at the end.” The data says: yes, 37% drop off in checkout – but 80% drop off between viewing a product and clicking “add to cart”. If you fixed only view_to_cart by 5 percentage points, you’d have more orders than from eliminating checkout abandonment entirely.
That doesn’t mean checkout can be ignored. 6,883 abandoned checkouts in 90 days is a real leak – and worth investigating separately. But starting with checkout without understanding what happens earlier is treating the symptom instead of the cause.
A magnifying glass on checkout – the per-channel breakdown
The overall numbers are a starting point. The real diagnosis begins when you break the funnel down by channel. Where checkout really “falls apart” – and for whom – is completely different from what the average suggests.
Each channel has a different weakness profile. The colors show: weakest vs best in each column.
| Channel | Sessions | View→Cart | Cart→Checkout | Checkout→Txn | CR total |
|---|---|---|---|---|---|
| 1,303 | 57.0% | 85.8% | 53.5% | 18.3% | |
| Direct | 14,133 | 22.2% | 93.4% | 67.1% | 10.5% |
| Cross-network (PMAX) | 36,848 | 18.7% | 82.2% | 61.7% | 8.2% |
| Paid Search | 77,514 | 28.7% | 75.2% | 61.3% | 5.8% |
| Organic Search | 62,309 | 11.0% | 75.2% | 59.1% | 2.0% |
| Referral | 4,549 | 10.2% | 76.5% | 60.3% | 5.0% |
Three findings that aren’t obvious:
1. Email has the worst checkout completion (53.5%) despite the highest overall conversion (18.3%). Users from email arrive with huge intent – 57% add a product to the cart. But then nearly half drop off in checkout. Something stops them on the home stretch: the form, no preferred payment method, a required account.
2. Direct has cart_to_checkout of 93.4% – almost no cart abandonment. Someone who already knows the brand and comes in directly almost always completes the cart. Here checkout_to_transaction is the highest too (67.1%). Loyal customers convert better at every stage.
3. Organic Search: 62 thousand sessions, only 2% CR. The largest traffic channel is by far the weakest at converting – and the problem is in view_to_cart (11%), not checkout. Organic users browse but don’t buy. That’s a signal to check: do the product pages match the intent of the queries they come from?
The per-channel breakdown changes how you think about optimization. Instead of “let’s improve checkout”, a more precise question appears: let’s improve checkout for Email users – because they have the highest purchase intent and at the same time the biggest problem at the last step. That’s a concrete, actionable hypothesis.
The Email paradox
The highest overall conversion (18.3%), but the worst checkout completion (53.5%). The email user is ready to buy – something stops them in the process itself. Check: does checkout require creating an account, how many fields the form has, which payment methods are available. There’s a concrete loss here.
When to raise the alarm
Four signals in the session_funnel data that call for action – not observation.
How often to check session_funnel
The funnel changes more slowly than campaigns – but it does change. It’s worth having a rhythm.
The main takeaway
Checkout isn’t the only point where you lose customers.
From 206,960 sessions to 11,551 transactions – along the way 94.4% of visitors drop off. Most of them drop off before checkout: 80% of people who view a product don’t add it to the cart. Checkout takes another 37%.
The per-channel breakdown says even more. Organic Search drops off at products (11% view_to_cart). Email drops off at completion (53.5% checkout_to_transaction) – despite the highest purchase intent of all channels. Each channel needs a different diagnosis and a different action.
That’s exactly what session_funnel is for: not to say “you have a conversion problem”, but to say which step, for whom and why.
DataOrganizer · MCP
DataOrganizer MCP connects an LLM with your GA4 – the full funnel with a per-channel breakdown in one question.