Practical guides on websites, online stores, and getting found on Google — written for business owners, not developers.
Tell us about your business today — get a free quote within 24 hours. No pressure, no confusing tech talk.
Tell us about your business today — get a free quote within 24 hours. No pressure, no confusing tech talk.
Ten design mistakes that quietly lose you customers, what causes each one, and how to fix it without a rebuild.
Your website talks to more potential customers than you ever will. It works nights and weekends, and it handles all the people who would never pick up the phone to call you in the first place. So why isn't it closing anything?
The average business website converts somewhere between 2% and 3.4% of its visitors, according to Contentsquare's 2026 benchmark, which is built on 99 billion sessions. Turn that number around and it means roughly 97 out of every 100 people who land on your site leave without doing a single thing.
Here's the part that surprises most owners. Those 97 didn't leave because the site looked bad. They left because they had to think too hard. What does this company actually do? Can they help me? What's this going to cost? How do I even get hold of someone? Every question you leave sitting there costs you a few seconds, and a few more visitors.
Psychologists call that cognitive load. Your visitor only has so much mental effort to spend, and your website is spending it in the wrong places.
Which is how a site can look genuinely beautiful and still convert at close to nothing. I've opened homepages that clearly cost real money and still couldn't tell you, after thirty seconds, what the business sold.
The ten below are the ones I run into most. The good news is that almost none of them need a rebuild — most are a copy change, a button moved, or something expensive you delete.
Most business owners try to fix this with more traffic. They hire an SEO, run some ads, and wait for the numbers to move. That order, SEO before web design, feels sensible enough. Get the visitors first, worry about the site later.
But traffic was never the problem. If 100 people visit and 2 convert, doubling your traffic just means 196 leave instead of 98. You're pouring more water into a bucket that still has a hole in it. SEO decides how many people arrive; design decides what happens to them after they do.
So where does the hole come from? Mostly from making design decisions on how a site looks rather than how it works. The video background wins the meeting because it's impressive. The animated menu wins because it's clever. Nobody in the room asks what either one costs the visitor, because the visitor isn't in the room.
That cost turns up in the same four places nearly every time:
Pages that load slowly
Navigation that confuses instead of guides
Messaging that never plainly says what you sell
A contact option buried three clicks deep
And there's one more thing that rarely gets said out loud. Your website gets reviewed on a large desktop monitor, over fast office WiFi, by people who already know exactly what your business does. Your visitor has none of those advantages. They're on a phone, on mobile data, and until about four seconds ago they had never heard of you.
They'll give you seconds, not minutes. If they can't work out what you offer or how to buy it in that time, they leave, and you never find out they were there.
Go and open your homepage. Read the first line.
If it starts with "We are" or "Our company," you've already lost a few of them.
Most hero sections get written to introduce the business. "We are a full-service digital agency delivering innovative solutions since 2015." It reads like a handshake at a networking event, which is fine at a networking event.
But nobody lands on your website to learn about you. They land because something is broken and they're working out whether you're the person who fixes it. You get about five seconds to answer that, and "since 2015" isn't an answer.
On kickoff calls I ask the owner to read their own hero section out loud to me. It usually ends the argument about whether it needs changing faster than any amount of explaining does.
The fix:
Lead with their problem, not your history. The first line should describe the situation your visitor is in right now.
Cut "we" and "our" from the headline. Swap in "you" and "your." If the headline still makes sense, keep it. If it falls apart, it was never about the customer.
Say what changes. Not what you do — what's different for them once you're done.
Same business, rewritten:
Before: "We are a full-service digital agency delivering innovative solutions since 2015."
After: "Your website gets traffic. It doesn't get enquiries. Let's fix the second part."
The rest of your homepage has other jobs to do, but none of them matter much if the hero fails.
That 3D scroll effect on your homepage cost you money twice. Once when you paid for it, and again every time it makes somebody wait.
Animation isn't really the villain here. Weight is. A single 10-second background video loop usually runs somewhere between 3 and 8MB. On office WiFi that's invisible. On mobile data in a car park it's a blank screen with a spinner on it, and your visitor is already tapping back.
Google measures this directly: Largest Contentful Paint should happen within 2.5 seconds, and Interaction to Next Paint should stay under 200 milliseconds. Heavy hero media and stacked animation libraries push both in the wrong direction, and those numbers feed your rankings, so this is a design decision that arrives with an SEO bill attached.
Then there's scroll-hijacking, where the page takes over your scroll wheel and moves at its own pace. It breaks Ctrl+F, it loses your place when you hit back, and it makes anyone in a hurry feel trapped on a ride they didn't buy a ticket for.
We set a page weight budget before design starts, on every build. One of ours had to work over satellite internet on a vessel, which is about as hostile as conditions get, but the discipline pays off on ordinary sites too.
The fix:
Decide the weight budget first, not last. Under 500KB is realistic for a business site. Anything over 1MB didn't happen by accident; somebody chose it.
Swap video backgrounds for a still frame. Pull one frame out of the video and use that instead. Almost nobody notices, and you get several megabytes back.
Animate on scroll, but never take over the scroll. Fade things in if you like. Just don't change how fast the page moves.
If you want to check yours, turn WiFi off, open the site on mobile data and count. That's the version your customer is actually looking at.
Most mobile sites were never really designed. They're the desktop layout squeezed until it fits.
You can spot one instantly: text shrinks to microscopic, buttons sit two millimetres apart, and the visitor goes to tap "Call" and hits "Cancel" instead.
Screen size is the easy part, and any decent theme handles it on its own. The hard part is physical. Fingers are wide, imprecise, and usually attached to someone who's walking.
The gap shows up in the numbers too. Contentsquare's 2026 benchmark puts desktop conversion at 3.4% and mobile at 2%, so desktop converts about 75% higher. Some of that is just how people browse. A lot of it isn't.
Apple puts the minimum tap target at 44×44 points, and Google's Material guidelines say 48×48dp. Shrink a desktop button down and it lands well under both.
The fix:
Design the mobile layout first, then expand it. Do it the other way round and mobile ends up as leftovers.
Give every tap target 44px minimum, plus breathing room. Two links sitting side by side need a gap, not a pixel.
Keep the important buttons in thumb reach. Bottom third of the screen, not the top-right corner.
Then test it properly: one hand, standing up, phone in your weaker hand. If you can't contact yourself in ten seconds, neither can they.
There's more to this than tap targets. Mobile-friendly design covers the layout side.
You know the ones. Four people in a bright office laughing at a laptop, a handshake in front of a window, a headset smiling at nothing in particular.
Your visitor has already seen those exact images on six other websites this month. A few will run a reverse image search. Most won't bother, because they can just feel it.
Here's the damage. Stock photography stopped reading as "professional" a long time ago, and now it reads as we couldn't show you anything real. On a small business site that lands somewhere between forgettable and suspicious.
Think about what a scam site looks like: polished, faceless, no evidence that actual humans work there. Your customer has been trained to spot exactly that. So when your site is all gloss and no proof, you trip the same alarm without meaning to.
That's the trap with looking corporate. Past a certain point, polish stops signalling credibility and starts signalling that something is being hidden.
When we brief photography for a build, we ask for the work rather than the office. The pictures a client is slightly embarrassed by are usually the ones worth publishing.
The fix:
Show the actual work. The job site, the van, the machine, the finished install. Slightly imperfect and real beats flawless and generic.
Put real faces on the team page. One decent phone photo of you at your desk will outperform a stock executive in a suit.
If you must use stock, use it as texture and never as proof. A background pattern is fine. A "happy client" photo of someone who has never been your client is not.
You don't need a photographer for this. You need thirty minutes, decent daylight, and a phone from the last five years.
Photos are only one part of how a site earns trust, but they're the part visitors judge first, before they've read a word.
"Leveraging disruptive paradigms to streamline enterprise experiences."
Read that out loud. Now picture saying it to a customer on the phone. You wouldn't, and nobody would.
Somewhere between the meeting room and the website, businesses start writing in a voice they'd never use in person. It's meant to sound credible, and it does the opposite, because the reader still can't work out what you sell.
There's a newer version of this problem too: copy that reads like it came straight out of a chatbot. Smooth, confident, completely empty. Your visitors read AI-written text all day now, and they've got good at recognising the rhythm of it.
The fix:
Write it the way you'd explain it to a friend. Then fix the grammar, but don't tidy the plainness out of it.
Swap abstract nouns for concrete ones. Not "solutions." Say the actual thing. Websites. Roof repairs. Payroll.
Cut any sentence that would work on a competitor's site. If it fits them too, it's telling your visitor nothing.
Read the finished page aloud before it goes live. Anything you stumble over is a sentence a visitor will stumble over too.
One more test: hand your homepage to someone outside your industry, give them ten seconds, then ask what you sell. If they hesitate, it's the copy that failed, not them.
"Great service, highly recommended!" — John S.
Who is John S.? What did he buy? What actually happened to his business?
Nobody believes an anonymous compliment. Worse, a wall of them reads like something you wrote yourself on a quiet Tuesday afternoon. And once a visitor suspects that one thing on your page is staged, they start re-reading everything else with the same suspicion.
There's a legal edge to this now, too. The FTC's Rule on the Use of Consumer Reviews and Testimonials took effect in October 2024. It bans fake or misrepresented testimonials, including AI-generated ones, and carries civil penalties of over $53,000 per violation. In December 2025 the FTC sent its first warning letters under that rule. So if you sell to US customers, invented testimonials aren't only a trust problem any more.
The more common situation, though, isn't fakery. It's an empty testimonials section on a business with plenty of happy customers who were simply never asked.
The fix:
Name the person properly. Full name, company, role, city. A link to their site or LinkedIn does more than a photo ever will.
Trade adjectives for outcomes. Not "great service," but something closer to "enquiries went from 2 a month to 11 in six weeks."
Put each one beside the claim it proves. A testimonial about fast turnaround belongs next to your turnaround promise, not buried in a carousel nobody scrolls.
Three specific, verifiable testimonials will beat twenty anonymous ones.
Nobody hides their contact details on purpose. It happens through structure.
The phone number sits in the footer, the form lives on /contact, and every other page quietly assumes the visitor will go looking for one or the other.
They won't. Buying intent is perishable. Someone decides to get in touch in a particular moment, and if that moment costs two clicks and a scroll, it passes.
On mobile it's worse. A phone number written as plain text isn't tappable at all, and a form means typing on a small screen while standing in a car park, then waiting two days for a reply.
The fix:
Put a tappable phone number in the header of every page. Not the footer, and not only the contact page.
Offer the channel they already use. For plenty of buyers that's WhatsApp or a direct call rather than a form.
End every service page with a way to act. Someone who has just finished reading about your service is the warmest they will ever be.
Contact friction is only one reason enquiries dry up. Here are the others.
Twelve fields. Company size. Budget range. Phone number, required. All before you've said a single word back to them.
But field count isn't really the villain here.
Plenty of businesses genuinely need the detail. A builder quoting a job, an agency scoping a project, a supplier checking specs; guessing wastes everyone's time, including the customer's. Asking for less isn't automatically better.
The problem is asking for all of it at once, upfront, with no explanation. That's the moment a form stops feeling like a conversation and starts feeling like a loan application.
The fix:
Break it into steps, easiest first. Two short screens feel lighter than one long one, even with identical fields. Add a progress bar so people can see it ends.
Say why, right next to the field. "A rough budget helps us send relevant examples, and a range is fine." One line removes most of the resistance.
Mark optional fields as optional. When everything is required, everything feels like an interrogation.
The working rule is to ask for whatever you need in order to send a useful first reply. The rest can come in the second message, once they're already talking to you.
There's more to it than field count. Here's what actually makes a contact form convert.
Get a Quote. Book a Call. Download the Guide. See Pricing. Join the Newsletter.
All on the same screen, all the same size, all the same colour.
Every extra option adds a decision, and decisions cost effort. Hick's Law puts it plainly: the more choices you offer, the longer someone takes to choose. Offer five equally loud ones and a good number of visitors take the easiest option available to them, which is leaving.
To be fair, having more than one CTA isn't the mistake. Giving them all equal weight is. When nothing on the page looks more important than anything else, the visitor ends up working out your priorities for you, which is a job they didn't sign up for.
The fix:
Pick one primary action per page, before design starts. What is this page actually for? Everything else drops to a text link or an outline button.
Repeat that one CTA instead of adding new ones. Three chances to click "Get a Quote" beats one chance each at three different things.
Match the ask to the temperature. A first-time visitor won't book a call. Give them something smaller and save the call for the page where they're ready for it.
Placement matters as much as count. Here's where CTAs actually belong on a page.
Nine items across the top, each one opening a dropdown, and a couple of those dropdowns have dropdowns of their own.
Menus grow this way because they follow the business rather than the visitor. "Solutions" exists because someone once ran a solutions department. "Resources" is where pages go when nobody could agree where else to put them.
Your visitor doesn't know your org chart. They arrived looking for one specific thing and now they're reading a directory.
On mobile it compounds, because all of that collapses into a hamburger icon and every extra level is one more tap between them and the answer.
The fix:
Cap the top level at five or six items. If everything sits in the menu, nothing is prioritised. The rest belongs in the footer, where people expect to go hunting anyway.
Name items after what visitors want, not what you call them internally. "Web Design," not "Solutions." "Prices," not "Engagement Models."
Test it on someone outside your industry. Give them a task, something like "find out what this costs," then watch where they click first and say nothing at all.
Here's the simple check: if a visitor can't reach your main service in one click from any page, the menu is working against you.
More on this in common navigation mistakes.
Most of these are cheap to fix. The expensive mistake is fixing them in the wrong order: new visuals on a page that still doesn't say what you sell, or a faster site that still has nowhere to click.
So we run every build and redesign through the same sequence.
Work out what the business actually sells, and who's buying it. Before a single layout decision. This is what fixes the hero and the copy, mistakes #1 and #5.
Decide what each page is for. One job, one primary action. That settles the CTA and navigation problems before they exist.
Write the content before the design. Not after. Design built around real words doesn't need lorem-ipsum boxes filled in later, and it doesn't end up with a beautiful section nobody has anything to put in.
Build to a weight budget. Page weight and tap targets get decided at build time. Retrofitting speed onto a finished site costs several times as much.
Collect the proof deliberately. Real photos, real testimonials. This is the step everyone skips, and then wonders why the trust section looks thin.
Measure after launch. Thirty days later we look at what people actually did, rather than what we assumed they'd do.
Our most recent redesign is a straightforward example. The homepage opened with a company introduction and little else, so a first-time visitor couldn't tell what was being sold. We rebuilt the hero around what the customer was looking for, and brought the products and services onto the homepage instead of leaving them a click away. We added an FAQ section covering the questions the owner was answering by email every week.
Then the unglamorous half: full descriptions written for all 250+ products, on-page SEO worked through page by page, and their local search presence set up properly. And because the site carried no reviews at all, we ran an email campaign to his existing customers asking for one.
Where it landed: the site now sits at position 4 in the local pack for its main service area, and it's the source Google cites in the AI Overview for their core query. It has brought in more than 15 quote requests. In marine calibration, jobs like those start at around $5,000.
That review campaign wasn't design work. But an empty testimonials section is rarely a design problem — it's usually just that nobody ever asked.
If your site has three or four of these problems, you probably don't need a rebuild. You need someone to fix them in the right order. That's the work we do.
How do I know whether design is my problem, or I just don't have enough traffic?
Check your analytics. If people are arriving and leaving without ever visiting a second page, more traffic won't help; you'll just lose more of them, faster. If almost nobody is arriving at all, that's a visibility problem to solve first.
Do I need a full redesign, or can I fix these one at a time?
Most of the ten are standalone fixes: a rewritten headline, a button moved, a video deleted. A full rebuild only earns its cost when the structure itself is wrong, meaning the pages don't match what people actually search for, or nothing can be edited without calling a developer.
Which of these matters most?
The hero and the contact path. If a visitor can't tell what you sell, nothing else on the page gets a chance. And if they can't reach you in the moment they decide to, everything else you fixed was wasted effort.
How long until fixes show up in enquiries?
Copy and CTA changes tend to show within weeks, because the traffic is already arriving and you're simply keeping more of it. Anything that depends on search visibility takes longer, usually a few months.
CEO
UI/UX Designer & Founder at AppDeel | We Design & Develop SaaS, eCommerce & Business Websites with Next.js, Shopify & Webflow
Read more posts by Azizul Hakim →Your site gets visitors but the form stays empty. These 7 reasons explain why people who could contact you don't and exactly what to change.
How to work out how long your site has actually been compromised, and why that answer decides everything else.
Google described our client's homepage as "Slide 1." Here is how design decisions quietly break SEO, and why website architecture should be planned first.
A breakdown of the fake Cloudflare verification scam hitting WordPress sites — how it hides from site owners, why security scanners miss it, and how we remove it.
Tell us about your business today — get a free quote within 24 hours. No pressure, no confusing tech talk.
