Cart Abandonment Is a Five Figure Problem Hiding in Your Analytics

Here is a number most shop owners never calculate, and it quietly ruins their year. Somewhere between the basket and the thank you page, a big slice of the people who already decided to buy from you walk away, and because it happens inside the checkout where nobody looks, it never shows up as a line you can be angry about. It is not a rounding error. For a shop doing real volume it is routinely a five figure sum every single month, sitting in plain sight in your analytics, waiting for someone to read the funnel from the bottom up instead of the top down. I want to show you how to find that number in your own data, how to size it with one honest formula, and why recovering even a thin slice of it is almost always worth more, and costs you far less, than buying another wave of traffic to paper over the leak. Because here is the thing nobody says out loud: you already paid to get those people to the basket. Losing them there is the most expensive way to lose a customer.

The leak you already paid for

Here is a number most shop owners never calculate, and it quietly ruins their year. Somewhere between the basket and the thank you page, a big slice of the people who already decided to buy from you walk away, and because it happens inside the checkout where nobody looks, it never shows up as a line you can be angry about. It does not arrive as a bad month. It arrives as a slightly disappointing every month, forever.

For a shop doing real volume that leak is routinely a five figure sum, and it is sitting in plain sight in your analytics right now, waiting for someone to read the funnel from the bottom up instead of the top down. The reason it hides is simple. Everyone stares at the top of the funnel, at traffic and clicks and cost per visit, because that is where the money goes out and where the agencies point. Almost nobody stares at the bottom, at the last few steps of the checkout, which is where the money that already walked in the door quietly walks back out.

I want to do three things with you. First, a quick self audit of what your web analytics is actually measuring, so you know which of the four boxes your leak lives in. Second, one honest formula that turns the leak into a euro figure you can put on a slide. Third, the bit that changes the argument in every meeting I have ever had about this: why recovering a thin slice of that leak is worth more, and costs you far less, than buying another wave of traffic to paper over it. Because you already paid to get those people to the basket. Losing them there is the most expensive way there is to lose a customer.

One more promise before we start. None of this needs a data scientist, a tool you cannot afford, or a quarter of workshops. Everything here runs on the analytics you already own and an afternoon of honest arithmetic. If you can read a funnel and multiply three numbers together, you can find your own leak tonight and size it by the weekend. That is the whole point of the exercise, to make an invisible loss visible enough that you cannot un see it, and then to price it so plainly that the decision about what to do next makes itself.

What your analytics is actually measuring

Before you can fix the leak you have to know what you are looking at, and most dashboards are a soup of numbers with no map. The cleanest map I know comes from BI, Analytics and Data Science, the 2018 textbook by Ramesh Sharda and colleagues, which sorts everything web analytics does into four plain categories. Read them as a self audit. Score yourself honestly on each, because the box you neglect is almost always the box the money is hiding in.

The first is usability. How do people actually behave on the site once they are on it. Where do they hesitate, where do they rage click, which step makes them reach for the back button. This is the box the checkout leak lives in, and it is almost always the box shops measure least, because it is the box that never sends you an invoice or a congratulations email. Nobody gets a prize for watching where a customer stalls on the payment screen, so almost nobody watches.

The second is traffic sources. Where visitors come from, which channel, which campaign, which keyword. This is the box shops measure most, because it is the box that sends invoices. Every ad platform on earth is desperate to show you this one in loving colour, because this is the box where you hand them money. It is not that traffic data is useless. It is that it soaks up all the attention while the real leak sits two boxes over, unmeasured and unloved.

The third is visitor profiles. Who these people are, new against returning, device, geography, the rough shape of the audience. Genuinely useful, and usually half ignored, sitting in a tab nobody opens between campaigns. It matters more than shops think, because the device split alone often tells you your checkout is failing on phones and passing on desktops.

The fourth is conversion. Whether the visit ends in the thing you wanted, a sale, a signup, a booking, and how much of the traffic gets there. This is the box the money hides in, and it is the box most owners glance at as a single top line percentage and never open up into its actual steps. That single percentage is the villain of this whole piece, because one number can never show you a gap, and the gap is where all the money is.

A quick scorecard for the four boxes

Do not just read the four categories, mark yourself against them. Here is the scorecard I hand people, and most fill it in with a very lopsided pattern the moment they are honest.

BoxWhat it answersHow often shops check itWhere the money actually is
UsabilityHow people behave step by stepRarelyHigh, this is the checkout leak
Traffic sourcesWhere visitors came fromConstantlyLow, it is a cost, not a cure
Visitor profilesWho the visitors areSometimesMedium, useful for targeting
ConversionDid the visit end in a saleOnly the top lineHigh, the leak lives in the steps

Here is the self audit in one question. Of those four, which do you actually look at every week. If the honest answer is traffic sources, and only the top line of conversion, you are in the majority, and you are also almost certainly leaving a five figure sum in the usability and conversion boxes untouched. Those two are the ones that pay you back, and they are the two nobody rotates onto. That is the first thing worth sitting with.

Reading your own funnel from the bottom up

Let me make it concrete with a shop I know well in shape if not in detail. Round numbers, one month, so the arithmetic stays honest and you can drop your own figures in tonight.

StageVisitorsDrop off from previous
Landing40,000none
Product viewed18,00055 percent
Add to basket8,00056 percent
Checkout started3,60055 percent
Purchase complete1,76051 percent

Read that table the way everyone reads it first, top down. Forty thousand people arrive, one thousand seven hundred and sixty buy, so the shop converts at about 4.4 percent, which is a perfectly normal number and nobody panics. Job done, close the tab.

Now read it bottom up, which almost nobody does. Eight thousand people liked something enough to put it in a basket. That is not a browser. That is a person leaning towards their wallet. Of those eight thousand, one thousand seven hundred and sixty came out the other side. The rest, more than six thousand people who had already chosen a product, did not finish.

The completion rate from basket to purchase deserves its own small formula, because naming a thing is what makes it fixable.

Completion rate=PurchasesBaskets=1,7608,000=22%\text{Completion rate} = \frac{\text{Purchases}}{\text{Baskets}} = \frac{1{,}760}{8{,}000} = 22\%

Twenty two out of every hundred people who reached for their wallet actually paid. The other seventy eight walked, most of them inside the last two steps, the steps you built, on the page you control. That is the leak. It is not at the top of the funnel where you spend all your attention and all your ad budget. It is at the bottom, among the warmest people you will ever have, and it is the one part of the whole journey that costs you nothing extra to improve because the traffic is already there.

Two ways to read the same funnel

The reason the leak hides is that the two honest ways to read a funnel tell completely different stories, and almost everyone only ever tells the first one.

ReadingWhat you seeWhat you concludeWhat you do next
Top down4.4 percent conversion, looks normalNothing is wrongBuy more traffic
Bottom up78 percent of baskets abandonedThe checkout is bleedingFix the checkout

Same numbers. Opposite conclusions. Opposite budgets. The top down read makes the checkout invisible and sends you to the ad platform. The bottom up read makes the checkout the most valuable square metre in the whole business. Which story you tell decides where your next euro goes, and most shops tell the one that costs them the most. The aha here is not that abandonment exists, everyone knows it exists. It is that the exact same table, unchanged, either hides the loss or screams it, depending only on which end you start reading from.

Putting a number on the leak

Vague leaks do not get fixed. Numbered ones do. So here is the one formula I put in front of an owner, and it is deliberately simple enough to trust.

Recovered=Baskets×Δ(completion rate)×AOV\text{Recovered} = \text{Baskets} \times \Delta(\text{completion rate}) \times \text{AOV}

In words, the extra revenue you recover equals the number of baskets, times the improvement in your basket to purchase completion rate, times your average order value. Three numbers you already have.

Let me run our shop through it. Baskets, we said, are 8,000 in the month. Average order value, call it 80 euros. And suppose we do the unglamorous work on the checkout and lift completion from 22 percent to 25 percent. That is a three percentage point lift, Δ\Delta equals 0.03. Not a miracle. A better checkout, fewer forced steps, a guest option, the payment methods people expect. Three points is a normal result, not a fantasy.

Recovered=8,000×0.03×80=19,200\text{Recovered} = 8{,}000 \times 0.03 \times 80 = 19{,}200

Nineteen thousand two hundred euros. In one month. From a three point change on a page you already own, applied to traffic you already bought. Over a year that same three points is roughly two hundred and thirty thousand euros.

Annual=19,200×12=230,400\text{Annual} = 19{,}200 \times 12 = 230{,}400

And notice we did not touch the top of the funnel at all. We did not buy a single extra visitor. We just stopped tipping so many of the warm ones out at the last step. If your instinct is that three points sounds small, that is exactly the trap. Three points on completion sounds tiny because it is a small looking percentage. Nineteen thousand two hundred euros a month does not sound tiny at all. Small percentages on warm traffic are large amounts of money. That is the second thing worth sitting with.

Find your own number in the table

Your baskets and your average order value are not mine, so do not borrow my nineteen thousand. Here is what one month of recovered revenue looks like across a range of shops, all at the same modest three point completion lift, so you can find the row nearest yours.

Baskets per monthAOV 40AOV 80AOV 120
2,0002,4004,8007,200
4,0004,8009,60014,400
8,0009,60019,20028,800
12,00014,40028,80043,200

Every cell is just baskets times 0.03 times AOV. Find your basket count down the side, your average order value across the top, and the number where they meet is your recovered revenue every single month from one dull checkout fix. Then double it in your head if you think you can win six points instead of three, which on a genuinely bad checkout is not greedy at all. The point of the table is simple. On any real volume this is a five figure number, and it is the same five figure number month after month, which is what makes it worth so much more attention than it gets.

The maths that should change your budget

Now the part that changes the argument. An owner hearing all this quite reasonably says, fine, but I could also just buy more traffic and make the same nineteen thousand. Could you. Let us actually check, because the answer is the whole point of this piece.

First, what is a visitor worth to us today. Our shop makes 1,760 sales at 80 euros, so about 140,800 euros of revenue from 40,000 visitors.

Revenue per visitor=1,760×8040,000=3.52\text{Revenue per visitor} = \frac{1{,}760 \times 80}{40{,}000} = 3.52

That is 3.52 euros of revenue per visitor at the current, leaky completion rate. To make an extra 19,200 euros purely by adding traffic, and leaving the checkout exactly as it is, you need

Extra visitors=19,2003.52≈5,450\text{Extra visitors} = \frac{19{,}200}{3.52} \approx 5{,}450

about 5,450 extra visitors every month. That is a fourteen percent lift in traffic, month after month, and it is not free. If those visitors come from paid channels at, say, 0.90 euros a click, that is very roughly 4,900 euros of new ad spend, recurring, every single month, just to stand still against a leak you never fixed.

Traffic route against checkout route, side by side

Put the two routes next to each other and the choice stops being a matter of taste.

QuestionBuy more trafficFix the checkout
Extra monthly revenue19,20019,200
What it costsAbout 4,900 every month, foreverOne piece of work, once
What it demands14 percent more visitors, permanentlyA better last two steps
When you stop payingThe revenue stops tooThe revenue keeps coming
Effect on all other trafficNoneEvery visitor now converts higher

The traffic route is a subscription you pay forever. The checkout route is a repair you pay once. And it is better than a straight comparison, because the two do not just compete, they stack. Fix the checkout and every visitor you buy from now on converts at the higher rate too. The completion rate is a multiplier that sits underneath all your traffic, bought and free, today and next year. Improving it lifts the yield of every euro you will ever spend at the top of the funnel. Buying traffic over a broken checkout improves nothing underneath. You are pouring more water into a bucket instead of patching the hole, and the ad platforms are more than happy to keep selling you water. That is the third thing worth sitting with, and it is the one that should move your budget.

The whole thing in one picture

Here is why the leak hides, drawn out. The mistake is optimising the wide top of the funnel and ignoring the narrow bottom, when the bottom is where the warm money is.

The solid line down the middle is the funnel everyone measures. The dotted line off the basket is the one nobody prices, and it is the expensive one, because everyone on that dotted line was a person you had already paid to bring in and who had already chosen what they wanted. They are not cold traffic. They are the warmest money in the building, leaving by a side door you built yourself.

A worked mini case, smaller shop, same lesson

Big round numbers can feel like they only apply to big shops, so here is a smaller one, still illustrative, so you can see the argument does not need scale to bite. Picture a homeware shop doing 12,000 visitors a month, 3,000 baskets, an average order value of 45 euros, and a basket to purchase completion of just 18 percent.

StageVisitors
Landing12,000
Add to basket3,000
Purchase complete540

Five hundred and forty sales from three thousand baskets. Now the same dull work lifts completion from 18 to 21 percent, three points again.

Recovered=3,000×0.03×45=4,050\text{Recovered} = 3{,}000 \times 0.03 \times 45 = 4{,}050

Four thousand and fifty euros a month, call it forty eight thousand a year, from a shop a fraction of the size of the first one. And the traffic comparison is even less flattering to the buy more route. A visitor here is worth

Revenue per visitor=540×4512,000=2.025\text{Revenue per visitor} = \frac{540 \times 45}{12{,}000} = 2.025

so to make that same 4,050 euros with traffic alone you would need about 2,000 extra visitors a month, a sixteen percent lift, bought and paid for every month, against a fix you could ship once. Different shop, different numbers, identical shape. The leak is not a big shop problem. It is a checkout problem, and every shop with a checkout has one.

The checkout fix pattern

So what does fixing it actually look like. The pattern is well worn, and one of the clearest public examples is the retailer Lotte com, whose team studied their own checkout behaviour and found a large share of people abandoning inside the payment and login steps rather than at the product. The lesson was not a clever trick. It was that the checkout itself, the forced account creation, the redundant steps, the friction right at the moment of paying, was doing the damage, and that removing friction there recovered a meaningful chunk of otherwise lost orders. I will not quote their figures at you, because your shop is not their shop, but the shape of the finding travels everywhere.

The moves that tend to matter are dull, which is exactly why they get skipped.

The frictionThe dull fixWhy it moves completion
Forced account creationOffer a guest checkoutNobody has to sign up to give you money
Too many steps and fieldsCut to the minimum the order needsLess to fill in, fewer places to quit
Payment methods hidden lateShow expected methods up frontPeople relax when they see their method
Surprise total at the endKeep delivery in the total the whole wayNo nasty shock on the final screen
Fiddly on a phoneMake it work with one thumbThat is where most baskets now live

None of that is glamorous. All of it moves the completion rate, and the completion rate, as we just saw, is worth five figures a month on real volume. Crucially, you find which of these is your problem by looking, not guessing. This is the usability box from the self audit, the one nobody measures. Watch where in the checkout people drop off, step by step, and the leak tells you where it is. The shop above lost most of its people between checkout started and purchase complete, so that is where the work goes. Yours may leak a step earlier. The data knows. Read it bottom up.

The common mistakes that keep the leak open

Over the years I have watched the same handful of mistakes keep the same leak open in shop after shop. None of them feel like mistakes in the moment, which is exactly why they last.

The first is reading conversion as a single number. If the only figure you ever see is 4.4 percent, the leak is mathematically invisible to you. One number cannot show a gap, and the gap is the whole story.

The second is optimising the ads while the checkout stays broken. It feels productive because there is always another audience to test, but you are polishing the top of a funnel that leaks at the bottom, and every improvement up top just pours faster into the same hole.

The third is forcing account creation before payment. It is the single most common cause of abandonment I see, and it is imposed for the shop's convenience, not the customer's. People came to buy a thing, not to start a relationship with your database.

The fourth is hiding delivery cost until the final screen. Nothing kills a warm buyer faster than a total that jumps at the last moment. If the number is going to appear, let it appear early, while they still feel in control.

The fifth is treating mobile as an afterthought. Most baskets are built on a phone now, and a checkout that needs two hands and a mouse quietly discards the majority of your buyers before they ever reach the payment step.

The sixth, and the quietest, is guessing where the drop off is instead of watching it. Opinions about the checkout are plentiful and usually wrong. The step level numbers are right, and they are already in your analytics waiting to be read.

The seventh is renting a checkout you are not allowed to change, which turns every fix above from a decision into a request. More on that in a moment, because it is the one that traps owners hardest.

How you actually do this without a data team

You do not need anything exotic to start. You need three things wired together, and you can have all three inside a fortnight.

First, a funnel you can actually read step by step, not just a single conversion percentage. You need to see basket, checkout started, and purchase complete as separate numbers, because the gap between two of them is your leak and the top line hides it. Most analytics tools will do this the moment you define the steps. If yours cannot, that is a reason to move, because you cannot fix what you refuse to measure.

Second, a checkout you are actually allowed to change. This is the quiet catch. If you are renting a locked down platform, you may not be permitted to remove that forced account step or reorder those fields, which means the single most valuable repair in your whole business is off limits to you. It is one of the main reasons I steer owners towards an ecommerce platform they own and control rather than a rented box that decides for you where your customers are allowed to leave.

Third, the willingness to read the funnel from the bottom up, every week, next to the traffic numbers, not instead of them. Traffic sources tell you what you are spending. The bottom of the funnel tells you what you are wasting. You need both on the same screen.

A four week plan you can actually run

If you want this to be a project rather than a good intention, give it four weeks and one owner. Here is the plan I hand people, and it fits around a normal working month.

WeekThe jobWhat you end with
1Define the funnel steps and split conversion into basket, checkout, purchaseThree numbers instead of one
2Size the leak with the formula, work out revenue per visitor and the traffic you would need insteadOne slide, one euro figure
3Watch the step level drop off and pick the single worst stepThe one place to work
4Ship the dullest fix for that step and measure the deltaA real, banked completion lift

Notice what is not in there. No rebrand, no new ad agency, no three month discovery. One person, four weeks, the analytics you already have, and a single dull fix aimed at the single worst step. That is the whole method, and it is deliberately small enough that nobody can say they did not have time.

Here is the decision, drawn out, once you can see the step level numbers.

The loop at the bottom matters more than any single fix. You are not looking for one heroic redesign. You are looking for the worst step, you flatten it, you bank the delta, and then the next worst step becomes the biggest, so you go again. A checkout gets better the way a floor gets clean, one honest pass at a time, and each pass pays for the next.

If you want a hand finding the leak in your own funnel and sizing it in euros before you spend another cent on traffic, that is squarely the kind of work I do, and it is what my ecommerce and conversion work is built around. Bring me your funnel and your average order value and we will put a number on the door your customers are walking out of.

The one number to leave with

If you take a single idea from this, make it this one. Before you buy another visitor, calculate what you are already losing between the basket and the thank you page, because it is almost certainly a five figure sum and it is almost certainly less expensive to fix than to out spend. Baskets, times a few points of completion, times average order value. Do that sum on your own numbers tonight. The people in that gap already chose you. They already cost you money to reach. Letting them leave at the last step is the most expensive habit in ecommerce, and it is the one nobody has put a price on. Put a price on it, and the whole argument about where your next euro should go answers itself.

If you suspect a five figure sum is walking out of your checkout every month, let us find it in your funnel and size it in euros before you spend another cent on traffic. Book a call and we will put a number on the door your customers are leaving by.

1%of every invoice goes to a UK charity you pick.

A donation, never sponsorship. You choose the cause at onboarding.

The story behind the pledge →

Stay ahead of your competition.

The latest innovative products and services, straight to your inbox before your competitors hear about them.

Get up to 5% off your first six months: 1% per topic you pick, the full 5% when you take everything. Limited offer · ends 31 December 2026.

New clients only. Terms apply.

* Up to 5% off your first six monthly invoices, new clients only. Full terms.

Questions about pricing, contracts or how we work together?

Read the FAQ