Run a Friction Audit: Finding the Silent Conversion Killers

Conversion

Most people who run a website assume that when conversions dip, the problem is the offer. The price is wrong, the discount is too small, the product is not compelling enough. So they discount harder, rewrite the headline, add another testimonial, and wait for the numbers to move. They rarely do, because the offer was almost never the real problem. In my experience running Identiti Design, the far more common culprit is friction: the accumulation of tiny frustrations, hesitations, and moments of confusion that quietly bleed intent out of a visitor before they ever reach the point of deciding. No single one of these moments feels fatal. A form field that asks for one thing too many, a page that takes a beat too long to load, a button whose label leaves you unsure what happens next. Individually they are forgettable. Stacked together across a journey, they are the reason someone who arrived ready to buy ends up closing the tab. A friction audit is the discipline of finding those silent killers on purpose, and it is one of the highest leverage things you can do for a site that is getting traffic but not converting.

Friction, Not the Offer, Is Usually the Problem

I want to be precise about what friction actually is, because the word gets used loosely. Friction is any resistance between the visitor’s intent and the action you want them to take. It is not the same as a weak offer. An offer can be genuinely excellent and still convert poorly because the path to accepting it is full of small taxes on attention, effort, and confidence. When we take on a new project, the very first assumption I try to kill is that the business is losing sales because customers do not want what is being sold. The traffic and the behaviour usually tell a different story. People arrive, they engage, they show clear signs of interest, and then they drop off at specific, repeatable places. Those places are friction points, and they are almost always fixable without touching the price or the product at all.

The reason friction is so dangerous is precisely that it is quiet. A broken checkout throws an error and someone complains. A confusing three step decision that makes a visitor feel slightly stupid produces no complaint at all. They simply leave, and they leave believing the fault was theirs or that they were not that interested after all. You never hear about it. Your analytics show a fall off, but not a reason. This is why I call these the silent conversion killers: they do not announce themselves, they do not generate support tickets, and they do not show up as bugs. They show up as a conversion rate that is lower than it should be, and a founder who cannot understand why great traffic is not turning into revenue.

There is a useful mental model here. Every visitor arrives with a finite budget of patience and motivation. Motivation is what you spent your marketing money to create. Friction is what spends it down. If the total friction across the journey exceeds the motivation the person arrived with, they leave before converting. Your job in an audit is not to add more motivation with louder claims and bigger discounts. It is to stop wasting the motivation you already paid for. That reframe changes everything about where you look and what you fix first.

Start by Defining the Primary Conversion Path

Before you audit anything, you have to decide what you are auditing. This sounds obvious and yet it is the step almost everyone skips. A website does dozens of things. It has a blog, a careers page, a newsletter signup, social links, product pages, a contact form, and half a dozen calls to action competing for attention. If you try to audit all of it at once, you will find a hundred small issues and no way to know which ones matter. So the first move in any friction audit is to name the single primary conversion path. What is the one action that, if a visitor completes it, represents real value to the business? For an ecommerce store it is the path from landing to completed purchase. For a service business it is the path from landing to a qualified enquiry. For a software product it is the path from landing to a started trial or a booked demo.

Once you have named the primary conversion, map the actual steps a person must take to complete it, in order, from their very first touch. Not the steps you imagine, the steps that genuinely exist. Land on a page, understand what is on offer, choose an option, add it to a cart, enter details, confirm, pay. Or land, read, click through to a form, fill it, submit, receive confirmation. Write these out as a literal sequence. This becomes the spine of your audit. Every friction point you find will sit on one of these steps, and knowing which step lets you weigh its importance. A tiny annoyance on a step that ninety percent of visitors must pass through matters far more than a glaring problem on a page hardly anyone reaches.

I also encourage clients to define what happens just before and just after the path, because friction leaks in at the seams. The ad or search result that brought someone in sets an expectation, and if the landing page does not immediately match that expectation, you have introduced friction before the journey even began. Likewise the moment after conversion, the confirmation, the next step, the reassurance, is where trust is either cemented or quietly damaged. Define the whole path, edge to edge, and you give yourself a map instead of a pile of disconnected observations.

Walk the Path as a First-Time User on a Real Device

Here is where the audit becomes real, and where it stops being a spreadsheet exercise. You have to walk the primary conversion path yourself, from start to finish, as though you had never seen the site before. This is much harder than it sounds, because you cannot un-know what you know. The founder knows where every button is, knows what the product does, knows the jargon. That knowledge is exactly what blinds them to the friction a stranger experiences. So you have to deliberately adopt the mindset of a first-time visitor who is mildly impatient, slightly distracted, and not especially forgiving. Ask at every step: do I understand what this is? Do I know what to do next? Do I trust this enough to continue? Is anything making me hesitate, squint, or think?

Do it on a real device, on a real mobile phone, on a normal connection, not on your fast office wifi with a large desktop monitor. The majority of traffic for most of the businesses we work with arrives on mobile, and the mobile experience is where friction hides in plain sight. Buttons that are easy to click with a mouse become fiddly with a thumb. Forms that look tidy on a wide screen become endless scrolls on a phone. Text that is comfortable at desktop size becomes a squint. Popups that are a minor irritation on desktop can make a mobile page nearly unusable. If you audit only on desktop, you will miss most of what is actually costing you conversions, because you will be auditing the experience that a minority of your visitors have.

The single most valuable thing you can do beyond walking it yourself is to watch someone else do it. Hand your phone to a person who fits your audience but has never used the site, give them a realistic task, and stay silent while they attempt it. Do not help. Do not explain. Just watch where they pause, where they backtrack, where they frown, where they ask a question out loud. Every one of those moments is a friction point you would never have found on your own, because you already knew the answer. Five minutes of watching a real person struggle teaches you more than an hour of staring at the interface alone.

The Four Kinds of Friction

To turn scattered observations into something you can act on, it helps to categorise friction rather than treat every problem as the same kind of thing. I use four buckets, and sorting each issue into one of them tells you a great deal about how to fix it. The first is cognitive load: friction that makes the visitor think harder than they should. Unclear language, too many options, cluttered layouts, ambiguous labels. The brain treats thinking as work, and every unit of unnecessary work is a small reason to give up. The second is effort, or interaction cost: friction that makes the visitor do more than they should. Long forms, extra clicks, repeated data entry, unnecessary steps. This is physical and procedural cost rather than mental cost, and it accumulates fast.

The third bucket is anxiety and trust friction: anything that makes the visitor feel unsure, exposed, or at risk. Uncertainty about cost, about what happens after they click, about whether the business is legitimate, about what will be done with their data. This kind of friction is emotional, and it is often the most powerful, because fear stops people more reliably than annoyance does. The fourth is technical and performance friction: slow pages, layout that shifts as it loads, broken elements, errors that are not handled gracefully. This bucket is the most objective and often the easiest to measure, and it is frequently the most neglected because it feels like an engineering concern rather than a conversion concern. It is very much a conversion concern.

The value of these categories is that each points toward a different kind of remedy. Cognitive load is solved by simplification and clarity. Effort is solved by removing steps and asking for less. Anxiety is solved by reassurance, transparency, and proof. Technical friction is solved by engineering and performance work. When you catalogue everything you found on your walkthrough and tag each item with its category, you stop seeing a chaotic list of complaints and start seeing patterns. Often a single site is dominated by one category, and that tells you where the leverage is. A site drowning in anxiety friction does not need a faster server. It needs to earn trust.

Cognitive Load: Value Proposition, Choices, and Navigation

The most expensive cognitive friction almost always sits at the top of the funnel, in the value proposition. Within a few seconds of landing, a visitor is asking three silent questions: what is this, is it for me, and why should I care. If your page does not answer all three quickly and clearly, the visitor has to work to figure it out, and many will not bother. I see this constantly: headlines that are clever instead of clear, that describe the company’s ambition rather than the customer’s outcome, that assume knowledge the first-time visitor does not have. The fix is rarely more words. It is sharper, plainer words that state the value in the customer’s own language. Reducing the mental effort of that first moment is one of the highest return changes you can make, and it sits inside the broader discipline of reducing cognitive load across the whole experience.

The second common cognitive killer is too many choices. There is a persistent belief that giving people more options is generous, that a longer menu, more plans, more products on the page, more links in the navigation, serves the customer. In practice the opposite is usually true. Every additional option adds to the time and effort of deciding, and beyond a certain point people cope with the difficulty of choosing by not choosing at all. This is the practical lesson of Hicks law and why fewer choices convert: the more paths you offer, the longer the decision takes and the more likely the visitor stalls. When we audit a page with a stalled conversion rate and find eight calls to action competing for attention, the fix is often to remove seven of them.

Navigation deserves its own attention because confusing navigation is cognitive friction that compounds. If a visitor cannot form a clear mental model of how the site is organised, every subsequent action costs more effort, because they are navigating a maze rather than following a path. Menus with vague or overlapping labels, categories that do not match how customers actually think, important pages buried three levels deep, all of these make people feel lost, and feeling lost is a strong reason to leave. Good navigation is invisible: the visitor always knows roughly where they are, what is available, and how to get back. When you walk your own path, notice every moment you have to stop and think about where to go next. Each of those is a cognitive tax you can remove.

Effort: Forms, Forced Accounts, and Hidden Costs

If cognitive load is the tax on thinking, interaction cost is the tax on doing, and nowhere is that tax higher than in forms. Forms are where intent goes to die. Every field is a small demand, and every unnecessary demand is a reason to abandon. I have seen enquiry forms that ask for a job title, a company size, a phone number, and a preferred contact time before a stranger has any reason to trust the business, and then the same business wonders why so few people complete it. The rule of thumb is brutal and simple: ask only for what you genuinely need to take the next step, and defer everything else. Whether a field earns its place, how you group and label fields, how you handle validation, all of it matters enormously, and it is worth treating form design for conversion as a discipline in its own right rather than an afterthought.

Forced account creation is a particular species of effort friction that deserves singling out, because it is so common and so costly. Requiring someone to create an account before they can buy or enquire inserts an entire secondary task in the middle of the primary one. The visitor came to do one thing and is now being asked to do a different thing first, to invent a password, to remember it, to hand over an identity, all as a precondition of giving you money. The number of purchases lost to this single decision is enormous, and the remedy is almost always to allow the action to proceed without an account and to offer account creation afterwards as an optional convenience. The business rarely needs the account as much as it thinks it does. It needs the conversion.

Hidden costs are effort friction wearing the costume of anxiety, and they are one of the most reliable killers in ecommerce. A visitor works their way through the whole path, decides to buy, reaches the final step, and only then discovers shipping charges, taxes, or fees that were never mentioned. The feeling this produces is a mix of irritation and distrust, and it is often enough to end the transaction on the spot, even when the total is still reasonable. The problem is not the cost itself. It is the surprise, and the sense of having been led along without full information. Be transparent about the full cost as early as possible. People will forgive a price they can see coming. They will not forgive one that ambushes them at the finish line.

Speed and the Technical Tax

Technical friction is the category most likely to be dismissed as somebody else’s job, and it is precisely because it lives in the gap between marketing and engineering that it so often goes unaddressed. Page speed is the clearest example. A page that takes several seconds to become usable is bleeding visitors before they have seen a single word of your carefully written copy. The frustrating thing is that this loss is completely invisible in the way most people look at their site, because the founder is loading the page from a fast connection on a device that has already cached everything. The first-time visitor on a mid-range phone on a patchy connection is having a very different experience, and that experience is where your conversions are being lost.

Speed is not only about the total load time. It is also about how the page behaves while it loads. A page where the layout jumps around as images and elements arrive, where a button moves just as someone goes to tap it, where content appears and then shifts, produces a jittery, unstable feeling that erodes confidence even when the raw numbers look acceptable. The same is true of interactions that lag, taps that do not seem to register, and elements that respond slowly to touch. All of this is technical friction, and all of it makes the site feel cheap and unreliable in a way that transfers directly onto how the visitor feels about the business behind it.

The reason I insist on treating performance as a conversion issue rather than an engineering nicety is that the cost is real money and the fix is usually well understood. Oversized images, heavy scripts that block rendering, third party tags that pile up over time, all of these accumulate quietly as a site grows, and no single addition feels significant. Then one day the site is sluggish and nobody can point to the moment it happened. During an audit, measure the real experience on a real device, identify the heaviest offenders, and treat every second you can shave off as directly protecting the motivation you paid to create. Fast is not a luxury. Fast is respect for the visitor’s time, and visitors reward it.

Anxiety: Trust Signals, Errors, and Unclear CTAs

Anxiety friction is the resistance created by doubt, and it strikes hardest at the moments of greatest commitment: handing over money, sharing personal details, taking an irreversible step. Weak trust signals leave the visitor to wonder whether the business is real, whether the product is any good, whether other people have taken this same step and been glad they did. Trust is built with specifics rather than slogans: genuine reviews, recognisable proof, clear contact information, transparent policies, and design that simply looks credible and cared for. A shabby or inconsistent interface is itself an anxiety signal, because people reasonably assume that a business careless with its website may be careless with their order or their data. Every reassurance you can place at the point of doubt reduces the friction of committing.

Poor error handling is a particularly cruel form of anxiety friction, because it strikes at the exact moment someone is trying hardest to succeed. A form that rejects an entry without saying why, an error message written in technical language, a submission that fails and loses everything the person typed, all of these turn a small stumble into a reason to abandon. Good error handling is generous: it tells the person exactly what went wrong, exactly how to fix it, right next to the field that caused it, and it never punishes them by discarding their work. When you walk your own path, deliberately make mistakes. Enter a bad email, skip a required field, submit too early. What happens next tells you whether your site helps people recover or quietly pushes them out the door.

Unclear calls to action are where cognitive load and anxiety meet, and they are astonishingly common. A button that says something vague leaves the visitor uncertain what will happen when they click, and uncertainty at the moment of action is a powerful brake. Will this add to my cart or buy immediately? Will this commit me or just show me more? Will I be charged now? A call to action should tell the person exactly what happens next and, ideally, what they get for taking it. The words on your buttons do real work, and getting them right is a craft worth studying through proven CTA copy formulas rather than defaulting to the same tired verbs everyone uses. Clarity here removes friction at the single most decisive moment on the whole path.

Prioritise Fixes by Impact Against Effort, Then Validate

By the end of a thorough walkthrough you will have a long list, and the danger at this point is doing everything at once or doing the easy things first regardless of whether they matter. The discipline that saves you is prioritisation by impact against effort. For each issue, estimate two things: how much it is likely affecting conversions, and how much work it will take to fix. Impact depends heavily on where the friction sits. A problem on a step that every visitor must cross carries far more weight than a problem on a page few people reach. Effort depends on whether the fix is a copy change, a layout adjustment, or a deep technical rebuild. Plot these against each other and the sequence becomes obvious.

The first things you do should be the high impact, low effort fixes, the ones sitting on the critical path that cost little to change. Clarifying a confusing headline, removing fields from a form, exposing a hidden cost, rewriting a vague button, enabling guest checkout. These often move the numbers quickly and cheaply, and the early wins build the appetite to tackle the harder work. The high impact, high effort items go next, planned properly rather than rushed. The low impact items, regardless of how easy they are, should wait, because time spent polishing things that do not sit on the primary conversion path is time not spent where the money is. A friction audit that does not end in a ranked plan is just a list of complaints.

The final and most important discipline is to validate rather than trust your own opinion. Everything you found in the walkthrough is a hypothesis, an informed one, but still a hypothesis. The fact that something felt like friction to you does not prove it is costing conversions, and the fact that a fix feels obviously better does not guarantee it will perform better. This is where you lean on real evidence, combining behavioural data, session recordings, heat maps, and structured user testing to confirm where people actually struggle, a toolkit worth understanding properly through solid conversion research methods. And rather than waiting for the primary conversion rate alone to tell you whether a change worked, watch the smaller signals along the path, because tracking micro conversions as leading indicators shows you much sooner whether a fix is helping people move forward or quietly making things worse. Where the traffic supports it, test changes properly before rolling them out, so that decisions are settled by evidence and not by whoever argues most confidently in the room.

A Checklist You Can Actually Run

If you want to run this yourself, here is the whole method in the order I would follow it. Start by naming your single primary conversion, the one action that matters most, and write out every step a visitor must take to complete it, from the ad or search result that brings them in to the confirmation that reassures them afterwards. Resist the urge to audit everything. Pick the path that carries the revenue and give it your full attention. That decision alone will make the rest of the audit sharp instead of scattered.

Next, walk that path yourself on a real phone, on a normal connection, pretending you have never seen the site before, and ask at every step whether you understand it, know what to do next, and trust it enough to continue. Then watch at least one real person from your audience attempt the same task without your help, and note every pause, backtrack, and frown. As you go, write down every point of resistance and tag each one as cognitive load, effort, anxiety, or technical friction. Pay particular attention to the usual suspects: an unclear value proposition, too many competing choices, forms that ask for too much, forced account creation, costs that appear late, confusing navigation, slow or unstable pages, unhelpful error handling, thin trust signals, and vague calls to action.

Finally, turn that catalogue into a plan. Score each item by likely impact and by effort to fix, do the high impact low effort changes on the critical path first, and schedule the heavier work deliberately. Before you trust any of your conclusions, validate them with behavioural data and testing, and watch the small forward steps to see whether your fixes are working sooner rather than later. Then do the whole thing again in a few months, because friction is not a problem you solve once. It creeps back in with every new feature, every added tag, every well meaning addition to the page. The businesses that convert well are not the ones with the cleverest offers. They are the ones that treat the removal of friction as ongoing work, and keep walking their own path with fresh eyes long after everyone else has stopped looking.