The site looks good. The team is pleased with it. It cost real money and took four months. And the enquiry form gets used about twice a quarter, usually by a recruiter.
This is the most common brief we get, and the diagnosis is almost never the one the client expects. They expect to be told the design is dated. Usually the design is fine. What is wrong is structural, and it is one of four things.
1. It describes the company instead of the decision
Most B2B sites are organised around the org chart. Home, About, Services, Case Studies, Contact. It is a filing system for information about you.
But nobody arrives wanting information about you. They arrive with a problem and a shortlist, and they are trying to answer one question: is this the right firm for my specific situation. A site organised as a filing cabinet makes them do the work of assembling that answer from parts.
The sites that convert are organised around the decision instead. They lead with the situation the reader is in, name the thing they are trying to get done, and only then explain how you do it. The information is often identical. The order is not.
2. The proof is unverifiable
Every B2B site claims results. Almost none of them can be checked.
"We increased their pipeline significantly." "Trusted by industry leaders." A logo wall with no story attached. A testimonial signed — Director, Manufacturing Company. A buyer has read the same sentences on six sites this week and has learned to skip them, which means the section of your site that was meant to build trust is being scrolled past.
Specific and checkable beats impressive and vague, every time, and it is not close. A named client, a real situation, what was actually built, what happened after. If the number is unremarkable, use it anyway — an unremarkable number that is obviously real does more work than a spectacular one that reads as marketing.
This is the single most common gap we find. Companies with genuinely good work describe it in language so cautious that it becomes indistinguishable from companies with none.
If a client will not be named, you can still be specific about the situation and the mechanism without identifying them. What you cannot do is retreat to adjectives.
3. There is one conversion path, and it is a big one
Most B2B sites offer exactly one way to raise a hand: a contact form, or "book a call".
That works for the small fraction of visitors who are ready to talk to a salesperson today. For everyone else — and it is most of them — there is nothing to do. They are interested but not ready, and the site offers them a binary: commit to a sales conversation, or leave. They leave, and you never knew they were there.
What is missing is a lower step. Something genuinely useful that does not require a meeting: a diagnostic, a checklist, a worked example, a piece of research about their market. It gives the not-yet-ready visitor a reason to identify themselves, and it gives you something to follow up about that is not "just checking in".
Both of ours work this way. The Lead Leak Audit is eleven questions and takes about two minutes; The Qualified 25 finds twenty-five real companies that should be buying what you sell. Neither requires anyone to talk to us. Both tell us a great deal about who is looking.
4. It is slow, and slow reads as small
Performance gets treated as a technical concern, filed with hosting. It is a trust signal.
A site that takes five seconds to become useful on a phone on a train tells the visitor something about how the company operates, and it is not flattering. They will not articulate it as "poor Core Web Vitals". They will just feel that you are a bit amateur, and go back to the results page.
The usual culprits are boring: enormous unoptimised hero images, a stack of marketing scripts each adding a tenth of a second, a font strategy that leaves text invisible while it loads, a video that autoplays and blocks everything else. Very little of it is load-bearing for the design. Most of it can go, and nobody will notice anything except that the site now feels fast.
What to change first
In order, because doing all four at once is how a redesign takes four months and changes nothing:
- Fix the proof. It is the cheapest change and usually the highest impact. One real, named, specific example beats a page of claims. This is copy work, not design work, and it can ship this week.
- Add a lower step. One useful thing that does not require a meeting. It does not have to be sophisticated to work.
- Reorder the homepage around the reader's decision rather than your structure. Often the same content, resequenced.
- Then look at speed, which is real but rarely the binding constraint on a site that is already only getting two enquiries a quarter.
The pattern underneath all four: a B2B site fails when it is built as a description of a company rather than as a working part of the sales process. The redesign that fixes it is frequently less visual than the client expects and more about structure, evidence and what happens after someone raises a hand.
How long should a B2B homepage be?
As long as it takes to answer the decision, which is usually longer than a designer wants and shorter than a founder wants. The length is a symptom; the sequence is the thing to get right.
Do we need case studies if our clients won't be named?
You need specifics, which is not quite the same. Describe the situation, what was built and what changed, without identifying the client. Anonymous and specific still works. Named and vague does not.
Is a redesign the answer?
Often not. If the design is competent, the four problems above are content and structure problems, and you can fix them on the site you have for a fraction of the cost.
How do we know which one is our problem?
Look at what happens after someone lands. High traffic and no enquiries points at proof or the missing lower step. Traffic that leaves in seconds points at speed or a homepage that does not name the visitor's situation.
More in B2B web design.