There's a specific kind of meeting I've sat in more times than I can count. The traffic numbers are up. Paid media is performing. The content team is publishing. Everyone did their job. And yet the form fills are flat, sales is quietly frustrated and somebody eventually says the thing everyone is thinking: "So where are the leads going?"
Usually, nobody suspects the website. The website looks fine. Somebody spent real money on it. It has a nice hero image and a clean layout and it loads reasonably fast. It looks, for lack of a better word, professional.
Looks fine and converts are two completely different things, and the gap between them is where a lot of marketing budget quietly disappears. You already paid to get that person to the page. You paid for the ad click, the content that ranked, the email that landed. Then they arrived, couldn't figure out what to do next and left. That's the most expensive kind of visitor there is.
After 15-plus years of designing sites and then watching how people actually use them, I've noticed the culprits are rarely dramatic. It's almost never one broken thing. It's a stack of small design decisions that each seemed reasonable in a review meeting and add up to a page nobody can navigate. Here are the five I see most often, and what I believe actually fixes them.
This is the number one conversion killer I run into, and it usually comes from good intentions.
Someone on the team says the certifications matter. Someone else says the new case study deserves a spot. Sales wants the product specs higher. Leadership wants the award badges visible. Every request is legitimate, so everything gets added and every element gets styled to feel important. Bold headline, bold subhead, bright button, bright badge, colored callout box, a banner announcing the webinar.
The result is a page with no hierarchy at all. When six things all compete for attention, the visitor's brain does the only thing it can. It stops trying and starts scrolling.
Visual hierarchy is not decoration. It's the instruction manual for a page. Size, weight, color and position tell a person what to look at first, second and third, as well as in what order the story should unfold. When a designer removes emphasis from four elements so one can dominate, that isn’t the designer being particular about aesthetics. The designer is making sure the one thing you actually want the visitor to do is the thing they see first.
The fix is simple, though it takes some discipline. Pick the one action that matters most on the page, then deliberately rank everything else below it. Secondary information can still be there, it just shouldn't be fighting for the same attention.
This is where a lot of well-intentioned content strategy runs directly into how humans actually behave.
Nielsen Norman Group has been studying this for nearly three decades, and the finding has barely budged. People don't read web pages, they scan them. Their eye tracking research found visitors typically read only about 20 to 28 percent of words on the page, and that the average page visit is short enough that thorough reading is essentially off the table.
Their most famous finding is the F-pattern, where a visitor's eyes sweep across the top, drop down and sweep across again a little shorter, then run down the left edge. Most people treat that as a rule to design around. But here's the part that gets left out, and it's the part I find most useful. NN/g's own follow-up research reframed the F-pattern as a warning sign rather than a law.
This pattern shows up when a page lacks structure, when headings are weak, when paragraphs are dense, when there's nothing for the eye to grab onto. Give people clear headings, short paragraphs, and real visual anchors, and the pattern breaks in your favor.
There are a few other predictable ways people move through a page, and each one tells you something about how that page is built. The ‘layer-cake’ pattern is what good structure produces: someone scans headings and subheads, then drops into the body copy only where a heading earns it. The ‘spotted’ pattern is someone hunting for a specific thing, like a price or a phone number, skipping past anything that doesn't look like it.
Spotted behavior on a page meant to persuade usually means the page is answering a question nobody asked. Chunking is what turns those habits in your favor: alternating section backgrounds, cards for different offerings, clear headings with space around them. Tabs and accordions are fine for detail someone will go hunting for, like specs or FAQs, but anything that has to persuade should stay out in the open.
Mobile deserves a mention too, since a single narrow column changes all of this. Nothing sits side by side and there's no left edge to run down, so the order you stack things becomes the hierarchy. A CTA that sits comfortably beside three certification badges on desktop can end up a full thumb-scroll from the content that motivated it, which is why you should always check stacking order before calling a page finished.
So when your beautifully written value proposition sits in a dense six-line paragraph in the middle of the page, it isn't being rejected: It's being skipped, and your visitor may not even realize it.
The fix is designing for the scan first. Front-load the important words. Make headings say something specific rather than clever. Break dense blocks apart. Assume your visitor will read four things on the page, and choose those four things on purpose.
Forms are where I see the most money left on the table, and the reasoning behind the mistake is always the same. Sales wants better qualified leads, so marketing adds fields. Job title, company size, budget range, timeline, how did you hear about us and a dropdown for the industry that somehow never includes the right option.
The data on this is pretty consistent. In a well-known Marketo test, trimming a form from nine fields down to five lifted conversion from 10 percent to 13.4 percent. HubSpot's analysis found that each additional field reduces conversion by roughly 4 percent on average. Every question you add is a small tollbooth, and some percentage of people turn around at each one.
Now, the honest nuance, because this is where a lot of agency advice gets lazy. Shorter is not automatically better. Research from CXL and others has found plenty of cases where reducing fields actually hurt results, usually because the removed fields were doing useful filtering, or because the form no longer felt substantial enough for a high-consideration purchase. A ten-field consultation request can absolutely outperform a three-field one if the offer justifies the effort.
It's not, ‘How few fields can we get away with?' but rather, 'Does what we're asking for match what we're offering right now?' A checklist download shouldn't need more than three or four fields, and asking for budget and timeline in exchange for a one-pager is asking too much for what's on the table. Ten fields on a site audit request is entirely reasonable, because every answer changes what you hand back. Match the ask to the moment, and let the sales conversation do the qualifying.
Progressive profiling is worth considering too, since it settles the argument between what sales wants to know and what a visitor is actually willing to fill out. Ask for a little now and more later. A first-time visitor trades name and email for a guide, then comes back for a webinar and the form only asks the one or two things you still need. Sales gets the same depth over time, and nobody faces a wall of endless form fields.
I audit a lot of sites; one of the most common things I find is a primary CTA that exists in a place no one is looking, styled in a way that doesn't read as clickable, saying something that doesn't tell you what happens next.
‘Learn More’ is the worst offender. It promises nothing and creates zero urgency. Compare that to "See How It Works" or "Get a Quote" or "Book a 15 Minute Call," where the visitor knows exactly what they're agreeing to. Ambiguity is friction, and friction is where people leave.
Placement is equally important: a single CTA parked at the very bottom of a long page assumes everyone scrolls to the end, which they don't. If a visitor reaches a moment of interest halfway down the page, that's the moment the next step should be available, not four sections later.
And then there's the styling problem. If your button uses the same color as your section dividers, or sits inside a busy area with three other colored elements, it stops registering as an action. Buttons need contrast, breathing room and consistency. When every CTA on your site looks and behaves the same way, visitors learn the pattern quickly. When each page invents a new one, they have to relearn it every time.
Buyer journey also matters, since one button can't serve everyone. Early on, a visitor is still deciding whether you're even relevant, so the right next step is low friction: read the guide, watch the short overview. Mid-consideration they can handle a bigger ask, like pricing or an assessment. The mistake I run into most isn't making a big ask early, it's making it the only option. ‘Book a 15 Minute Call’ in a landing page hero works fine when someone arrived from a high-intent search or an email that already did the convincing. It falls flat as the lone CTA on an awareness-stage page, where the honest next step is something smaller.
Whitespace is the first thing to get sacrificed in a review, usually with some version of "there's a lot of empty space here, can we fill it?"
I understand this instinct. Empty space feels like waste, especially when there's a long list of things the page is supposed to communicate. But whitespace isn't necessarily empty, it's can serve as the thing that tells a visitor where one idea stops and the next begins. It may be what makes a headline feel like a headline instead of just a slightly larger sentence. It's what makes a button look pressable. Take space away indiscriminately and every element bleeds into its neighbors, so the page reads as one dense wall of company information rather than a clear path to a decision.
There's also a credibility dimension that's easy to underestimate. Generous spacing can read as confident. Cramped, cluttered layouts can read as either dated or desperate, and in B2B, where a visitor is deciding whether to trust you with a significant purchase, that impression forms in seconds and quietly colors everything after it.
Density is not the same as substance. A page with fewer elements and more room around them almost always communicates more, not less.
Here's what makes conversion work worth doing. You're not trying to get more traffic: the traffic is already there. The goal is to stop losing these people, which means the math tends to move faster than anything else in marketing.
If your site converts at 1.5 percent and careful design work moves it to 2.5 percent, you didn't get a 1 percent improvement. You got roughly 65 percent more leads from the exact same traffic, ad spend and content investment. That's the part that tends to land in a leadership meeting.
The reason this work needs a designer's eye rather than a checklist is that the problems above interact with each other. Fixing your form doesn't help much if nobody scrolls far enough to see it. Improving your CTA copy doesn't matter if the button is invisible. Adding whitespace does nothing if the hierarchy still gives six elements equal weight. You have to look at the page as a whole system, in the order a real person experiences it.
This is the kind of problem we genuinely enjoy, partly because the answers are usually hiding in plain sight.
Our approach starts with looking at what's actually happening on your site rather than guessing. Where are people landing, where are they dropping off, what are they clicking and where does the path to a conversation break down? We use analytics, heatmapping and session data to find the leaks, and we lean on AI to move through that analysis faster than we could manually, spotting patterns across pages that might take days to surface by hand.
Then the design work begins, and this is where having brand and website expertise in the same building matters. A conversion fix that ignores your brand produces a page that converts slightly better and feels like it belongs to a different company. A brand refresh that ignores conversion produces something beautiful that doesn't perform. Doing both at once means the hierarchy, the messaging, the visuals and the path to contact all pull in the same direction As a result, the site ends up looking like you while working considerably harder for you.
That's the opportunity in front of most B2B companies right now. You've likely already invested in the traffic. The work of turning more of it into actual conversations is the highest leverage thing left on the table, and it usually costs a fraction of what you're spending to bring those visitors in.
If you take one thing from this, make it this exercise. Pull up your most important page, look at it for five seconds, then look away and ask yourself what you were supposed to do. If the answer isn't immediate and obvious, your visitors are having the same experience, except they don't work there and almost certainly aren't going to give you a second try.
You already paid for them to show up. The least the page can do is tell them where to go.
Kuno's Website Design and Brand Experience teams help B2B companies find and fix the gaps between traffic and conversion. If you're seeing solid visitor numbers and soft lead numbers, that's a conversation worth having.