Six signs your B2B website is costing you deals
Sites that lose deals usually look fine. They lose deals because a buyer couldn't work something out and didn't ask.
Most conversations about B2B website design turn into conversations about taste, which is a shame, because taste is rarely the thing losing you deals. Sites that lose deals usually look fine. They lose deals because a buyer couldn't work something out and didn't ask.
That's the failure mode worth worrying about. Nobody emails to say I didn't understand your homepage so I went with someone else. They just don't reply, and you record it as a slow quarter.
What follows are six signs we see repeatedly in B2B and SaaS companies between roughly seed and Series B. They're diagnostic rather than aesthetic — you can check most of them in an afternoon without a designer's opinion. Each one comes with the cheap fix and the real fix, because sometimes the cheap fix is genuinely enough and it's dishonest to pretend otherwise.
One framing before the list. Your website has two jobs now, and only one of them is the one it was built for. It has to convince a person, and it has to be legible to the machines summarising you to people who never visit. A site that fails the first fails slowly. A site that fails the second fails invisibly.
1. Your sales calls start with "so, what do you actually do?"
This is the cleanest signal there is, and it costs nothing to check. Listen back to five recent first calls. If the prospect opens by asking you to explain the product — a prospect who booked the call, from your website — then the site did not do its job. It generated a meeting and delegated the explaining.
That's survivable while you're taking every call yourself. It stops being survivable the moment you have a salesperson, or the moment the buyer is comparing you to two others and only one of them made itself obvious.
Cheap fix: rewrite the first screen to say what the product does in one sentence that a stranger could repeat correctly. Test it by sending it to someone outside your industry and asking them to explain it back.
Real fix: you probably have a positioning problem wearing a website costume. No amount of copy editing fixes a product nobody has decided who is for.
2. Your homepage describes a category, not a change
Open your homepage. Read the headline. Does it name a thing you are, or a difference you make?
"The AI-native platform for revenue teams" names a category. It's also indistinguishable from about four hundred other homepages, which is why SaaS homepage design has converged so hard that you can swap logos between competitors and nobody notices for a week.
Buyers don't purchase categories. They purchase a specific change in their working life, and they're scanning for the sentence that describes it. If your headline requires them to infer the change, most won't bother.
Cheap fix: write the headline as the outcome, then use the subhead for the category so search and skimmers still get it. Outcome first, classification second.
Real fix: find out what your best customers say the change was. Not what they say in testimonials — what they said in the first ten minutes of the call where they decided. That sentence is almost always better than anything written in a workshop.
3. Nothing on the site survives being quoted out of context
Here's a test that takes three minutes. Copy your homepage text into a document, strip the images, and read it. Then ask ChatGPT or Claude what your company does, using only that text.
A lot of B2B sites fail this badly, and the reason is usually a design decision rather than a writing one. A page made of a three-word headline, a product screenshot, a logo wall and an animated demo can be beautiful and still contain almost no extractable prose. There's nothing for a model to lift, and nothing for a prospect to paste into Slack when they're selling you internally to their boss.
That second one matters more than people think. In B2B, somebody who isn't on your website has to be convinced by somebody who was. Your site has to survive being forwarded.
Cheap fix: make sure every page contains at least one paragraph of plain prose that states the point completely, with no reliance on the image next to it.
Real fix: treat your site as a source document rather than an experience. The parts that get quoted are the parts that do the work.
4. Your pricing page says "contact us" and nothing else
There are businesses where this is correct. Enterprise deals with genuine variability, heavy implementation, negotiated everything. If that's you, fine.
For everyone else it's a tax you're paying without noticing. A buyer with a budget and three options open in three tabs will disqualify the one that won't indicate a range, because contacting you to find out costs them thirty minutes and a sales sequence they don't want. They're not being lazy. They're managing their own time.
And it's worse than it used to be, because that buyer is increasingly asking an AI to shortlist for them first. A page that says "contact us" gives the model nothing to work with, so you get left out of the comparison before a human ever sees you.
Cheap fix: publish a range and the shape of what drives it. "Most teams land between X and Y; it depends on seats and whether you need SSO." You've lost no negotiating room and saved everyone a call.
Real fix: decide who you're not for, and let the page say so. Disqualifying the wrong buyers on the site is cheaper than disqualifying them on call three.
5. Every page is a landing page, so none of them is
This one shows up as a symptom: your SaaS landing page count keeps growing, every campaign gets a new one, and none of them are noticeably better than the homepage.
What's usually happened is that pages are being produced to satisfy internal requests rather than to answer a specific question at a specific moment. So you end up with eleven pages that all say roughly the same thing at roughly the same altitude, and a buyer who can't tell which one to read.
The symptom to check for: if you can't say, in one sentence, who each page is for and what they should do next, it's decoration.
Cheap fix: pick the three pages that matter, make them excellent, and unlink the rest from your navigation. You don't have to delete them.
Real fix: build pages against questions rather than campaigns. One question, one page, one next step. It's a smaller site and it works harder.
6. You have no page for the thing that kills your deals
Every company has one. The objection that comes up on nearly every call and has to be handled live, every time. Security. Migration effort. "We already have something that half does this." Being too small. Being too new.
Almost nobody puts it on the website, because it feels like leading with weakness. So instead it gets handled thirty times a quarter by the person whose time is most expensive, and it goes unhandled entirely for every prospect who never got as far as a call.
Addressing it publicly does two things. It removes the objection from the deals you'd have won anyway, and it builds more credibility than any amount of confident copy, because a company that names its own hard question reads as a company that has thought about it.
Cheap fix: one honest page. The objection as the heading, your actual answer underneath, no spin.
Real fix: ask your sales calls what else is in there. There are usually three, not one, and the other two are more interesting.
Three of these don't need a redesign
Worth saying plainly, given who's writing this: signs 1, 4 and 6 are copy and decision problems. You can fix all three in a fortnight without touching the design, and if that's where you are, a redesign would be an expensive way to avoid the actual work.
Signs 3 and 5 are usually structural, which means they're a content and information architecture job — sometimes a rebuild, often not. Sign 2 is the one that tends to be genuinely upstream of everything else, and it isn't a website problem at all.
The honest rule: rebuild when the structure can't hold the thing you now need to say. Not when the site looks dated. Looking dated costs you far less than being unclear.
The website redesign checklist
If you want to run this properly, in order, and it's about half a day:
- Listen to five recent first calls. Count how many open with "what do you actually do?"
- Read your homepage headline aloud. Category, or change?
- Strip your homepage to text. Ask an AI what your company does. Note what it gets wrong
- Send that same text to someone outside your industry. Ask them to explain it back
- Open your pricing page. Could a buyer self-qualify in 30 seconds?
- List every page in your nav. Write one sentence per page: who it's for, what they do next. Delete the ones you can't
- Ask whoever takes sales calls for the three objections they handle most. Check whether any has a page
- Search your own category in ChatGPT and Perplexity. See whether you come up, and whether the description is right
Anything you couldn't complete is the list.
What this is really about
B2B website design gets discussed as a visual discipline and bought as one, which is why so many redesigns change how a company looks and not how it sells. The sites that work are the ones where somebody made hard decisions about who the product is for, what it changes, and who it isn't for — and then had the nerve to put those decisions on the page in plain language.
The design part matters. It just can't do the deciding for you.
If you go through the list above and most of it comes back clean, your website isn't your problem and you should go and fix something else. That's a genuinely good outcome and nobody in our industry ever says it.
We run this as a free 30-minute teardown: no deck, just the list and what we'd do about it. Book one here.
