A B2B purchase is rarely one person clicking buy. It is a director who found you, a VP who has to justify the spend, a technical lead looking for reasons to say no, and a procurement contact who wants a security document. Most B2B sites are built for none of them. They are built for the founder, who reviewed the mockups, liked the hero image, and never had to buy anything from the site. That gap is why traffic goes up and pipeline stays flat.
Good B2B website design serves a buying committee rather than one visitor. That means separate paths for the economic buyer, the technical evaluator and the end user, proof placed where objections appear, most content ungated, and a small set of pages that carry pipeline: solution pages, comparison pages and case studies.
Why do most B2B sites fail their buyers?
The usual failure is not ugliness. Plenty of handsome B2B sites produce nothing. The failure is that the site answers the company's favourite question, which is what we do, instead of the buyer's question, which is whether this reduces the specific risk I am carrying this quarter.
Three patterns show up in almost every audit we run. The homepage describes the company in abstractions that would fit any competitor. The site assumes a single reader who moves in a straight line from awareness to demo. And every piece of useful information sits behind a form, so the technical evaluator who was sent to check your integration list gives up and reports back that they could not find anything.
Long sales cycles make this worse, not better. If the average deal takes four to nine months, the site is not a closing tool. It is a reference document that gets opened repeatedly by different people, often forwarded internally, often read at 11pm before a meeting. Design for rereading and forwarding, not for a single dramatic first impression.
Before touching design, we ask the sales team for the last ten deals and the last ten losses, then read the email threads. The objections that repeat become the section headings on the solution pages. It is unglamorous research and it changes the sitemap more than any workshop does.
Who is actually on the buying committee?
Committee size scales with contract value. A $500 per month tool might involve two people. A $150,000 platform commitment can involve six or more, and each of them has a different definition of a good decision. The site does not need six versions of itself. It needs content that lets each role find their evidence quickly and hand it to the next person.
| Role | What they are worried about | What they need from the site | Where it belongs |
|---|---|---|---|
| Economic buyer (VP, C-level) | Return, risk, and how this looks if it fails | Business outcome framing, pricing signal, named proof from similar companies | Homepage, solution pages, case studies |
| Champion (director, manager) | Building an internal case they can defend | A shareable one-pager, ROI logic, comparison against alternatives | Comparison pages, downloadable summaries, pricing |
| Technical evaluator | Whether it works with what they already run | Integration list, API docs, architecture notes, limits | Docs, integrations page, technical resources |
| End user or team lead | Whether this makes their week harder | Product screenshots, workflow walkthrough, onboarding detail | Product pages, demo video, help centre |
| Procurement and legal | Contract terms, vendor risk | Security page, compliance status, standard terms, support SLA | Trust or security page, footer |
| Finance | Budget fit and billing mechanics | Pricing structure, contract length, invoicing options | Pricing page, FAQ |
Notice that only two of those roles want a demo. The rest want documents. A site that offers a demo button and nothing else is asking four people to do work they cannot do.
What belongs above the fold?
Four things, in this order of importance. A specific statement of what you do and who for, written so a competitor could not copy and paste it. A single sentence naming the outcome. Two calls to action, one high commitment and one low, because a first-time visitor and a returning evaluator need different doors. And a proof strip that is honest, whether that is client logos you actually have permission to use, an integration set, or a compliance badge you genuinely hold.
What does not belong: a rotating carousel, a stock photo of a boardroom, or a headline built from category jargon. If your hero could be pasted onto a competitor's site with no edits, it is doing nothing. We would rather see a plain sentence like software that helps regional insurance brokers cut quote turnaround from days to hours than any amount of polished vagueness.
How should proof be presented?
Proof fails when it is stored in one place. A case studies page collects dust because nobody browses case studies for fun. They read the case study you put in front of them at the exact moment they doubted you.
Break proof into pieces and distribute it. A one-line result under the relevant feature. A two-sentence quote next to the pricing table where the price objection lands. A full case study linked from the solution page for the industry it belongs to. Structure each case study the same way so the committee can compare them: situation, constraint, what changed, measurable result, timeline. If you cannot report a number, report a mechanism honestly and say the number is confidential. Inventing metrics is the fastest way to lose a technical evaluator, because they check.
Third-party signals carry more weight than anything you write about yourself. Review platforms, analyst mentions, customer conference talks and integration partner listings all count. So does published thinking that shows you understand the problem, which is where a deliberate content programme earns its budget rather than acting as filler.
Should you gate content or leave it open?
The honest answer is that gating works for a shrinking category of assets and hurts everywhere else. Gate something only when the asset is genuinely expensive to produce, when the person who wants it is close to buying, and when you have sales capacity to follow up quickly. Original benchmark research, a detailed calculator or a bespoke assessment can justify a form. A four-page overview PDF cannot.
Everything else should be open, indexable and easy to forward. Technical documentation, integration lists, security details, pricing logic and product walkthroughs all get read by people who will never fill in a form, and those people influence the decision. Search engines and answer engines cannot cite what they cannot reach either, which quietly removes you from the shortlist before a human ever visits.
- List every asset and mark who reads it: champion, evaluator, or buyer.
- Ungate anything an evaluator or end user needs to do their job.
- Keep gates only on assets with a genuine follow-up motion behind them.
- Replace removed gates with lighter conversion points, such as an email summary or a saved comparison.
- Track assisted pipeline for six months, not form fills for six weeks.
When a client insists on keeping a gate, we run the ungated version on a duplicate URL for a quarter and compare influenced pipeline rather than lead count. In our own projects the ungated version usually produces fewer leads and more qualified conversations, which is an argument worth having with data instead of opinion.
Which pages actually generate pipeline?
Very few. In most B2B accounts we look at, four page types produce the overwhelming majority of sourced opportunities, and the rest of the site exists to support them.
| Page type | Job it does | Signal it captures |
|---|---|---|
| Solution or use case pages | Matches your product to a named problem and industry | High intent, easy to qualify |
| Comparison and alternatives pages | Answers the shortlist question directly | Very late stage, often the last page before contact |
| Pricing page | Filters, and gives the champion numbers to work with | Strong intent even without a form fill |
| Case studies by industry | Removes the will this work for us objection | Mid to late stage, forwarded internally |
| Documentation and integrations | Clears technical veto | Rarely converts directly, blocks deals when missing |
Blog posts still matter, but treat them as the top of a system rather than the system itself. A post that ranks and links to nothing is a vanity asset. A post that ranks, answers the question fully and routes to the relevant solution page is a pipeline contributor. The same logic applies to paid traffic: if you are running demand campaigns into a homepage, you are paying for clicks that land on the least specific page you own.
How do you measure a site with a nine-month cycle?
Form fills are a weak proxy and last-click attribution is close to useless in B2B. Better to build a small set of measures that survive a long cycle. Track page groups against opportunity creation rather than individual sessions. Ask every inbound contact one open question about what they read before reaching out, and record the answer in the CRM. Watch returning-visitor depth on solution and pricing pages, because repeat visits from the same company usually mean an internal conversation is happening.
Then improve deliberately. Rewriting a solution page and watching what happens to opportunity quality over the next quarter is slower than a two-week test, but it is the honest timescale. Faster feedback is available on the mechanical layer, where form friction, load speed and mobile rendering can be fixed and measured quickly through ordinary conversion work without waiting on the sales cycle.
Frequently asked questions
What makes B2B website design different from B2C?
B2B design serves several people deciding together over months, while B2C serves one person deciding quickly. Content has to be easy to forward internally, proof has to be industry specific, technical detail should be open rather than gated, and success is measured against pipeline created rather than same session conversions.
Should a B2B website show pricing?
Usually yes, at least a structure. Hiding price completely forces your champion to guess, and their guess is often wrong in a way that ends the deal internally. If pricing varies, publish the model, the factors that move it and a realistic starting point rather than nothing at all.
How many pages does a B2B website need?
Fewer than most companies build. A focused site needs a homepage, one page per core solution or industry, a product page, pricing, case studies, a trust or security page and a resources hub. In our projects that is commonly 12 to 25 pages at launch, then expanded as sales reports recurring objections.
Is a demo request the right primary call to action?
It works for a champion who is already convinced, but it excludes the rest of the committee. Pair it with a lower commitment option: pricing details, a self serve product tour, or a technical overview. A single high commitment button collects few leads and loses evaluators silently.
How long does a B2B website redesign take?
For a mid market company, roughly 10 to 20 weeks from kickoff to launch is realistic. Discovery and messaging take three to five weeks, design four to six, and build four to eight, with content running in parallel. Slippage almost always comes from late internal content and approvals rather than development.
Do B2B blogs still generate leads?
They influence deals more reliably than they create them outright. A post that answers a genuine buying question and links to the matching solution page contributes to opportunities that get credited elsewhere. Volume driven posts rarely contribute. Judge them on assisted pipeline and on which pages come up in sales calls.
The takeaway
Treat the site as a document that gets forwarded rather than a brochure that gets admired. Map the committee, give each role something to take away, open up the technical detail your evaluators need, and concentrate effort on the handful of pages that carry intent. Then measure against opportunities created over a quarter, because a nine-month sales cycle will not respond to a two-week test.