UX is a conversion lever

Friction is anything that makes it harder than necessary for the visitor to reach the goal. It rarely sits in one place, it spreads through the flow. The most common sources are easy to recognise:

• Cognitive load and too many choices. A page that asks the visitor to think, compare or decide too much at once loses the person who only wanted to do one thing. Fewer and clearer choices lower the threshold.
• Poor form design. The form is often where most people drop off. Needless fields, unclear error messages and requirements that only surface when the visitor tries to submit make people leave right before the goal.
• Unclear copy and CTA. If the visitor does not understand what happens after the click, or what the next step is, they hesitate. A button that just says "Send" without context creates more hesitation than one that says what the visitor actually gets.
• A slow page. Waiting is friction. A page that loads sluggishly loses visitors before the content has even appeared, and the sense of lag lingers through the whole flow.
• Mobile as an afterthought. When the layout is built for desktop and then squeezed onto mobile, buttons get too small, forms too long and text too dense. A large share of the traffic then meets the worst version of the page.
• An unclear next step. The visitor should never have to hunt for what to do. When the next step is not obvious, the movement through the funnel stalls, and a stalled visitor rarely converts.

Baymard Institute↗ has, through extensive usability research, mapped how forms and checkout flows make visitors abandon a purchase, and their recurring conclusion is that most drop-off comes from avoidable friction, not from a lack of interest. Anyone who wants to find their own friction starts not in a checklist but in their own flow: where does the visitor hesitate, where does it get fiddly, where do they disappear?

Friction is rarely visible as a single fault, but it leaves traces in the data. Before you rebuild anything, it is those traces you should read, because they show where the visitor actually struggles, not where you think they do:

• Funnel drop-off. When a clear share of visitors disappears between two steps, the friction usually sits in the step they leave. The funnel points out where, even if it does not always say why.
• Form abandonment. Visitors who start filling in but never submit signal that something in the form is stopping them: one field too many, a requirement they did not understand, an error they could not fix.
• Rage clicks (repeated frustrated clicks). When visitors click hard or over and over on the same element, they are trying to make something work that does not. It is friction that shows up almost in real time.
• Short sessions and quick exits. Visitors who leave almost immediately rarely even got started. That often points to the wrong expectation, a slow page or an unclear first impression.

This is raw data, not conclusions. A high drop-off figure says where, but which hypothesis explains why, and what you test first, belongs in hypotheses and prioritisation. Reading the signals measured against the business, not against how the page looks, is the heart of conversion optimization: decision support, not status. And because the analysis rests on your own tracking, that data should be yours, so the lessons stay with you even if you change supplier: you own the data.

UX work for conversion tends to derail the same way: the focus lands on the wrong things. Five pitfalls keep coming back:

• Persuasion tricks before usability. Countdowns, popups and "only two left" get layered onto a flow that is still fiddly. It is repainting a door that jams. Remove the friction first, add the persuasion second.
• Redesign by taste instead of data. A new design because the old one "felt dated" swaps what you know for what you guess. Without knowing where visitors actually get stuck, a redesign risks moving the friction rather than removing it.
• Dark patterns. Tricking the visitor into a purchase, a subscription or a consent with deceptive design may lift a figure in the short term, but it damages trust and leads to regret, returns and churn. Avoid them: short-term conversion at the cost of a long-term relationship is a bad deal.
• Desktop-first thinking. When the design is made on a large screen and mobile becomes an adaptation after the fact, most visitors meet the worst version. Start from the smallest screen, and the flow holds all the way up.
• Accessibility as an afterthought. An interface that shuts out someone who sees poorly, navigates with a keyboard or uses a screen reader also shuts out conversion. Accessibility is moreover a legal requirement for many, but the line to regulation and WCAG belongs to accessibility in web development, not here.

Nielsen Norman Group's review of deceptive design↗ shows why the tricks cost more than they give: visitors who feel deceived do not come back. The common denominator in the pitfalls is that they try to win conversion without making it easier to convert. An experienced hand does in an hour what an inexperienced one takes four to do, often precisely because it knows which of these dead ends are not worth going down.

UX, conversion optimization and web development run into each other, and a lot of UX work stalls because the problem really sits in an adjacent discipline. A simple split helps:

• It is UX and CRO if the friction sits in the flow: too many steps, unclear copy, forms that misbehave, an unclear next step. It can be observed in visitor behaviour and improved in design and content.
• It is a web job if the foundation is technical. A page that loads too slowly is itself poor UX, but speed is fixed in the build, not in the copy. How page speed is measured and improved belongs to Core Web Vitals.
• It is accessibility if the interface shuts out visitors with different abilities. It overlaps with UX, but is also driven by WCAG and legal requirements, which accessibility in web development goes through.
• The statistics behind the decisions are a craft of their own. Whether a change actually had an effect, and when a result is worth trusting, belongs to A/B testing and significance, not here.

Google's guide to page experience↗ is a good way in to where UX meets tech, since speed, stability and mobile-friendliness are both ranking signals and sources of friction. The dividing line is rarely knife-sharp, but it is often the first question worth asking: does the problem sit in the flow, in the build or in the measurement? If you want someone to read your conversion flow and point out where the friction actually sits, see how Memorise works with conversion.