Where mobile loses the sale

What is mobile conversion rate optimization, and where does it pay?

Mobile conversion rate optimization is the work of raising the share of mobile sessions that complete a defined action, a purchase, a lead form or a signup, measured as conversions divided by mobile sessions. It pays better than the desktop equivalent for a blunt reason. Mobile carries roughly 75 to 77% of e-commerce site traffic but produces only about 59% of global e-commerce sales,⁶ so a single point of improvement lands on a much larger base.

That gap is the brief in one line. Attention moved to the phone. Revenue did not follow at the same rate. Before anyone argues about button colours, the useful question is where those mobile sessions stop, and the answer is almost always further down the funnel than the homepage.

What is a good mobile conversion rate in 2026?

There is no single number, and the published benchmarks disagree enough that treating any one of them as a target is a mistake. Industry aggregation puts mobile e-commerce conversion at 1.5 to 3% and desktop at 3.2 to 3.9%.² A separate device breakdown ranks tablet highest at 2.9%, desktop second, and puts the sector average at 2.4%.¹ Both are credible. They are counting different sets of sites.

The number that matters more than any of them is the trend. The mobile to desktop conversion gap has narrowed to roughly 0.2 percentage points, down from about 1.5 points five years ago.² Phones are no longer structurally worse at converting than laptops.

A wide mobile to desktop conversion gap is no longer evidence that phones convert badly. It is evidence that something specific in your mobile experience is broken, and it is usually in the checkout.

So set the benchmark internally. Measure your own mobile and desktop rates on the same definition and the same window, and manage the size of the gap between them. The published bands are useful only for a sanity check on whether you sit far outside normal.

Where does mobile actually lose the sale?

In the checkout. Checkout usability research built over 14 years, across 344 benchmarked sites and more than 30,000 reviewed checkout elements, puts the global average cart abandonment rate at 70.19%.³ The device split is the part mobile teams should read: mobile abandonment runs at roughly 78 to 85% against 66 to 74% on desktop, a gap of 10 to 15 points.⁴ Device level benchmarks place mobile near 80%, tablet near 74% and desktop near 67%, against that 70.2% all device average.³⁴

A 10 to 15 point abandonment gap on the device carrying three quarters of the traffic is the largest single recoverable number most e-commerce teams hold. It is also the one least likely to be fixed by a redesign, because the causes are mechanical rather than aesthetic.

That figure is routinely misread as a pricing problem. It is a sequencing problem. The shopper did not object to the total in principle, they objected to meeting it at step four of five. Showing shipping, tax and fees on the product page and in the cart, before the checkout starts, recovers most of that loss without a single discount and without touching margin.

How much does mobile page speed affect conversion?

Enough to be a revenue line rather than an engineering hygiene item. Real user data shows the average site loading in about 1.9 seconds on mobile, while 53% of mobile visitors abandon a site that takes longer than 3 seconds.⁷ The sensitivity is steeper than most teams assume. A 0.1 second improvement in load time raised conversions by 8.4% for retail sites and 10.1% for travel sites.⁸

The cleanest controlled evidence comes from Core Web Vitals work. Core Web Vitals is Google’s field measurement set covering loading (Largest Contentful Paint), responsiveness (Interaction to Next Paint) and visual stability (Cumulative Layout Shift). In an A/B test on the Rakuten 24 storefront, improving those metrics lifted conversion rate by 33.13% and revenue per visitor by 53.37%.⁵

The trap is measuring the wrong version of speed. Field data, collected from real visitors on real phones and real networks, is what tracks conversion. A lab score from a synthetic test on a fast machine and a fixed connection can pass while mid-range Android users wait. Teams that optimize until the lab score turns green, without ever checking their field data, frequently ship no commercial change at all.

Which mobile checkout optimization fixes move the number?

In rough order of return per hour of engineering time.

1. Show the landed cost early. Shipping, taxes and fees on the product page and in the cart. This addresses the largest stated abandonment cause³ and costs nothing in margin.
2. Turn on the native wallets. Enabling Apple Pay and Google Pay at checkout is associated with an 8 to 12 percentage point reduction in mobile abandonment.⁹ It removes the two hardest tasks on a phone, typing a card number and typing an address.
3. Collapse the steps. One click checkout is associated with roughly a 22% reduction in abandonment.⁹ Where a full one click flow is not available, the intent still holds: fewer screens between decision and confirmation.
4. Fix the keyboard. Correct input types, autocomplete attributes and validation that fires as a field is left rather than on submit. On a phone, the right keypad appearing for a postcode field is worth more than a new hero image.
5. Offer guest checkout by default. Forced account creation is a tax collected before any value is delivered. Offer the account after the order is placed, when the shopper has a reason to want one.

Every item on that list is checkout work, and none of it requires a rebrand. If the mobile funnel shows a large drop between cart and payment, that is where budget goes before acquisition does. Fixing the conversion step first raises the return on every channel feeding it, which is why we treat conversion work as part of a digital marketing programme rather than a separate project run by another team.

Is a mobile app better for conversion than a mobile site?

On the checkout metric, yes, and by a margin worth knowing. Mobile web checkouts abandon about 32% more often than native mobile app checkouts.⁹ The reasons are structural rather than aesthetic. Apps hold saved payment credentials and addresses, authenticate biometrically, and skip the cold start load entirely.

That does not make an app the right answer for most businesses. The honest reading is narrower: a competent mobile web checkout still leaves something on the table, and most of that gap is closable without shipping an app. Saved payment methods, biometric re-authentication through the browser payment APIs, carts that persist across sessions and an installable progressive web app recover a good share of the same advantage at a fraction of the cost. Build the app when retention and repeat purchase justify it, not to rescue a checkout that has never been optimized.

How do you run mobile conversion rate optimization without fooling yourself?

Five disciplines separate teams that move the number from teams that merely ship changes.

Segment everything by device

A blended conversion rate averages the mobile problem away. Every funnel report, test result and dashboard tile should split mobile from desktop by default, not on request.

Score against field data

Use real user Core Web Vitals as the pass or fail gate. Lab tools are for diagnosis and for reproducing a fault, not for declaring victory.

Rank fixes by exposure

The fix worth doing first is the one sitting on the step the most mobile sessions actually reach. Count sessions at each step before ranking anything by how bad the screen looks.

Test on the mobile segment alone

A change that helps desktop can hurt mobile, and a blended test result will hide it. Read every experiment by device before calling it.

Watch where the sessions start

A growing share of mobile arrivals now begin in an assistant or an AI answer rather than on a results page, which changes both the traffic mix and the intent behind it. That is its own discipline, covered in our note on optimizing a site for voice and answer engines, and it feeds the same funnel you are repairing.

None of this needs a large programme to begin. Pull last month’s funnel by device, find the step where mobile drops hardest against desktop, and ship cost transparency and wallet payments against it. For a second read on where a mobile funnel is leaking, talk to us.

Written by the group's editorial team with the practice leads who run these builds. Reviewed before publish. Spotted an error? Tell us and we will fix it.