Why Your Website Is Not Converting on Mobile: The Five Failures

Three neighbouring diagnoses are assumed here rather than repeated. If nobody is arriving at all, the problem is upstream and is covered in why your website gets no traffic. If nobody converts on any device, the failure is the offer or the routing rather than the screen size, and that full diagnosis — including how to verify your enquiry form actually delivers — is in my website gets traffic but no leads. Page performance has its own three-clock breakdown in why your website is slow, and none of those mechanics are re-explained below.

This piece covers one specific case: desktop visitors enquire and phone visitors do not. That is the most common version of the problem, because phones are usually the majority of the traffic, and it is the version that hides best behind a site-wide average.

Get the Number That Proves It First

Before changing anything, split the conversion rate by device category and then by landing page. Both cuts are necessary. Device alone tells you phones are underperforming; page-level tells you whether it is the whole site or two templates, and that distinction is the difference between a week of work and an afternoon.

The reason this goes unnoticed for years is arithmetic rather than negligence. If phones are seventy percent of sessions and convert at a quarter of the desktop rate, the blended figure reads as unremarkable and nothing in it points at a device. Owners look at a mediocre overall number, conclude the market is soft, and buy more traffic — which then arrives mostly on phones.

Then do the thing no dashboard can do for you. Pick up your own phone, leave the office wifi, find your site the way a customer would, and complete an enquiry from beginning to end. Not a glance at the homepage. The whole task, on cellular, with one hand, outdoors if you can. Almost nobody who paid for a website has ever done this, and it usually identifies the failure in under two minutes.

What Actually Changes About a Person on a Phone

Most mobile advice treats the phone as a small desktop, which is why most mobile advice stops at layout. Five conditions change simultaneously, and every failure below descends from one of them. The visitor is standing rather than seated. They have one hand available, and the working area is a thumb arc, not a whole screen. They are interrupted and expect to be interrupted again within a minute. Their connection is worse than yours. And critically, they have a phone in their hand, which means calling is the path of least resistance and typing is the expensive option.

That last condition reverses the desktop assumption entirely. On a laptop, a form is the convenient choice and a phone call is a small undertaking. On a phone it is the other way around, and a site that only offers the form is asking for the harder thing from the person who has the easier one available.

Failure One: The Ask Is Typing When the Visitor Would Rather Tap

This is the biggest one and the cheapest to fix. A phone visitor in a service category usually wants to talk to somebody now. If your only route to contact is a form with six fields, a large share of them leave and call whoever is next in the results.

Two specific defects cause most of the damage. The first is a phone number rendered as plain text rather than a tap-to-call link, so tapping it does nothing and the visitor has to memorise ten digits, leave the page and re-enter them. That single omission is worth real money and takes minutes to correct. The second is a number that exists but sits below the services, the about section and a testimonial carousel — present, unreachable, and therefore functionally absent.

Offer both routes and put the tappable one within the first screen. Keep the form for people who prefer to write and for after hours, and if calls are answered during limited hours, say which hours on the page rather than letting someone discover it by ringing out.

Failure Two: Tap Targets Sized for a Mouse

A cursor is a single pixel and a thumb is roughly a fingertip wide. Controls that are comfortable with a mouse become a game of chance with a thumb, and the cost of a mis-tap is not one wasted attempt — it is often a navigation to the wrong page, a back button, and a decision about whether this is worth continuing.

Three patterns cause almost all of it. Links sitting immediately adjacent to each other in a stacked menu, so the wrong one is as likely as the right one. Small icon-only controls, particularly close buttons on overlays, which is the worst possible place to put an unreliable target. And text links buried inside a paragraph as the only route to something important, when the important route should be a button the thumb can find without aiming. Where the control sits also matters: the reachable zone on a modern phone held in one hand is the lower two thirds of the screen, and a primary action parked in the top corner asks for a grip change.

Failure Three: The Furniture Stack Eating the Screen

This one accumulates rather than being designed, which is why nobody owns it. Each piece was added by somebody reasonable for a defensible reason. A cookie or consent notice. A sticky header with the logo. A promotional bar. A chat bubble in the corner. A newsletter prompt that appears after eight seconds. On a laptop these are mild furniture. On a phone they can leave a strip of actual content in the middle of the screen.

Count them on your own phone and measure the honest result: how much vertical space remains for the thing the visitor came for. Then be ruthless about sequence, because two of these overlapping is a trap — a consent notice under a chat bubble means the visitor cannot dismiss either, and their only remaining option is to leave. Nothing should appear before the visitor has had the chance to read anything, and anything that appears must be closable with a target a thumb can hit on the first try.

Failure Four: The First Screen Says Nothing

On a desktop, a large photograph with a slogan across it is a reasonable opening, because the services, the location and the contact details are all visible in the same view. Narrow that same layout to a phone and the photograph becomes the entire first screen. The slogan survives. Everything that identifies the business does not.

So the visitor's opening three questions all go unanswered: what does this business do, does it operate where I am, and how do I reach it. Those are the only questions that matter in the first five seconds, and a phone visitor will not scroll to find them from a cold arrival. The fix is unfashionable and it works: plain words near the top saying what you do and where you do it, then the tappable contact route, then the photography. Businesses that make this one change often see the mobile gap narrow before anything else is touched.

Failure Five: The Form Fights the Keyboard

A form that works on a laptop can be close to unusable on a phone for reasons that never show up in a browser preview, because the software keyboard takes up half the screen the moment a field is focused and the layout has to survive that.

The recurring defects are specific. Fields that do not declare their input type, so somebody entering a phone number gets a full alphabetic keyboard and somebody entering an email has to reach for the at symbol. Text sized small enough that the browser zooms in when the field is focused, leaving the visitor scrolled sideways and lost. A submit button that ends up underneath the keyboard with no way to scroll to it. Validation that clears what was already typed when one field is rejected — on a phone, re-entering four fields is where people quit. And a field count that made sense to whoever designed the intake process rather than to the person standing outside with a problem: every field you do not act on is a reason for somebody to stop.

The Connection, Which Is a Different Problem From a Slow Site

Performance belongs in this article only in its mobile-specific form, and the general mechanics are in the speed post linked at the top. The mobile-specific part is that your visitor is not on your connection. They are on cellular, possibly indoors, possibly in a building that eats signal, on a device several generations behind the one you tested with.

What that changes is the failure mode. A page that is merely sluggish on a good connection often does not finish loading at all on a poor one, and a visitor who never sees the page cannot be counted as a bounce with a cause. This is also where layout shift does its damage: content arriving in stages means a thumb aiming at a button presses whatever moved into that position instead, and on an unreliable connection those stages are seconds apart rather than milliseconds.

What a Normal Gap Looks Like, and When Yours Is Not

Not every gap is a defect and chasing parity is a waste of money. Phone sessions genuinely contain more casual browsing, more people checking an address, and more arrivals that were never going to end in an enquiry. A mobile conversion rate somewhere around half to two thirds of desktop is ordinary in most service categories and is not worth an intervention.

The thresholds worth acting on are narrower than people assume. A mobile rate under about a quarter of desktop means an interface failure rather than an audience difference. A mobile rate of effectively zero on a page that converts on desktop means something is outright broken — a form that does not submit, a button that cannot be reached, an overlay that cannot be dismissed — and that is a same-day fix rather than an optimisation project. And a gap that appeared suddenly points at a specific change: a new consent notice, an added chat widget, a theme update, a fresh tracking script.

The Order to Work In

Sequence matters more than budget here, because the cheapest items are usually the largest. Make the phone number a tap-to-call link and put it on the first screen. Count the sticky elements and remove or delay everything that is not load-bearing. Put a plain statement of what you do and where you operate above the photography. Cut the form to the fields you genuinely act on and set the input types. Only then look at anything that requires a designer.

Those first four are edits to an existing site, not arguments for replacing one. A rebuild earns its place when the template cannot absorb the changes without breaking, or when the same fixes have been made and undone by successive updates — the broader test for that is in signs you need a new website. When it is the right call, mobile behaviour should be specified at the start rather than checked at the end, which is how we approach AI website building: built with Claude Code against the phone case first, because for most of these businesses the phone is most of the traffic.

The Arrival That Is Growing Fastest Lands on a Phone

One more reason this is worth fixing now rather than next year. A rising share of visitors do not arrive from a results page at all. They asked an assistant, received two or three recommendations, and tapped one — and that tap lands on a phone, from someone who already has a specific question and expects the page to confirm an answer they were given seconds ago.

Those visitors are the least patient traffic on the internet, because they were handed a shortlist and can return to it instantly. A page that opens with a photograph and a slogan converts them no better than a bad advert. This is why we treat the citation work and the landing experience as one job rather than two: the specific, structured content that gets a business named in answer engine optimization is the same content that answers the question on arrival, and the search side feeds the same pages. Getting named and then losing the visit is the most expensive way to succeed.