Nearly half of B2B websites leave a buyer with nothing on at least one question

Ask a B2B website twenty questions a buyer would plausibly ask, and most of the time you get something back: a full answer, a half answer, or a polite "talk to sales". Sometimes you get nothing at all. No page, no paragraph, no form. In the Storylane Buyability Index, 242 of 529 B2B websites (46%) left at least one question with nothing. I call these dead ends, and I think they cost more than their small share of questions suggests.

A quick word on the data, because it shapes what you can conclude. The Buyability Index scanned 529 B2B websites between Sep 8 and Oct 20, 2026, and checked each one against 20 buyer questions, 10,580 in all. The questions are generated per company (what a buyer would plausibly ask that business), not logged from real buyers. The sites were chosen by the people who ran the scans, so this is not a random sample of the web. "Nothing found" means the scan found no public page that answers the question. The answer may exist in a PDF, behind a login or in a rep's head. To a buyer reading the website, that is the same thing.

What is a dead end on a B2B website?

A dead end is a buyer question the website doesn't answer, half answer, or route anywhere. It is different from a half answer, which gives the buyer something to work with, and different from a sales gate, which at least tells the buyer where the answer lives.

Across the 10,580 questions, the split looked like this: 58.6% fully answered, 33.1% partly answered, 2.0% gated or sent to sales, and 6.3% dead ends. Six percent sounds small. But each site was asked only 20 questions, and nearly half of them hit at least one dead end. A real buyer asks more than 20.

Are dead ends a deliberate sales tactic?

The obvious explanation is that companies stay quiet on purpose. Keep the contract terms vague, the thinking goes, and the buyer has to book a call.

The numbers don't support that. If silence were a strategy, you would expect to see it written down as one: "contact us for terms", "ask your account manager". Those deliberate gates and deflections came to 208 questions. Dead ends came to 662, more than three times as many.

Pricing is the one topic where gating is a real habit: 158 pricing questions were gated or sent to sales, against 124 that hit a dead end. Take pricing out and the gap gets wide. Outside pricing, the Buyability Index found 538 dead ends and 50 sales gates. For every question a site deliberately routes to sales, it leaves about eleven with nothing.

A gate is a decision someone made. A dead end is a question nobody thought about. Most of the silence on B2B websites is the second kind.

Which buyer questions hit dead ends most often?

Dead ends are rare where marketing is comfortable and common everywhere else. Product questions dead-end 2.5% of the time. Pricing, comparison, implementation and support questions dead-end at roughly three times that rate.

Source: the Storylane Buyability Index, 529 sites, Sep 8 to Oct 20, 2026. A further 174 questions fell outside these topics (vendor risk, exit terms, who owns the work) and dead-ended 15.5% of the time.

Look inside the topics and one theme stands out. Questions about the contract (minimum term, cancellation, renewal, month to month) dead-ended 12.2% of the time, about five times the rate for product questions. Only 39% of those 721 contract questions got a full answer. Questions about single sign-on and SCIM dead-ended 10.1% of the time.

The pattern holds by buyer role. Questions a procurement or finance buyer would ask dead-ended 8.6% of the time; questions a technical buyer would ask, 4.8%. Questions that need a high level of assurance dead-ended 7.6% of the time, against 4.6% for questions a self-serve buyer could settle alone.

Why do websites go silent on these questions?

I have a theory, and it is about who owns the page.

A B2B website is written by marketing, about the product. So product questions get answered: someone's job depends on it. The questions that dead-end are about the relationship after the signature. How long is the contract. What happens to my data if we leave. What is your uptime commitment. How long until my team is live. Those answers belong to finance, legal, customer success and services. None of those teams owns a page on the website, and marketing doesn't feel it can write their answers.

I call these ownerless questions. They fall between teams, so nobody writes them down, and the website says nothing. That is why the silence clusters where it does: not on hard questions, but on questions that belong to someone other than the person editing the site.

This also explains why dead ends are a site habit rather than bad luck. 287 sites (54%) had no dead ends at all. 85 sites (16%) had three or more, and those 85 sites held 69% of all the dead ends in the index. Their median Buyability score was 54 out of 100, against 83 for sites with none. Some companies have found a way to get ownerless questions answered. Most of the rest have the same handful of gaps, again and again.

Why is a dead end worse than a half answer?

A half answer gives a buyer something to forward, quote or follow up on. If your site says "implementation usually takes a few weeks depending on data volume", the buyer has a range and a reason. They can ask a sharper question on the call. We have written about the assurance gap between easy and hard buyer questions, and half answers are a big part of it. They are a problem worth fixing, but they keep the buyer on your page.

A dead end sends the buyer somewhere else. Usually that is a rival's website or an AI assistant. And an assistant asked "what is the minimum contract for this vendor?" will answer from whatever it can find: a review site, a forum thread, a competitor's comparison page. When you leave the question empty, someone else fills it. That is the real cost, and it is why I'd rather a website could answer a buyer who arrived from ChatGPT on the questions it currently skips.

The strongest objection is that some answers really do depend on the deal. Contract terms get negotiated; SLAs vary by tier. That's true, and it is not a reason to say nothing. "Annual by default, with month to month on smaller plans" is an answer. So is "terms depend on volume; here is what usually changes". Buyers don't need your final number. They need to know the question has an answer and roughly what shape it takes.

How do you find and fix dead ends on your own site?

You don't need a big project. You need a list and a few owners.

1. Ask the ownerless questions yourself. Search your own site, as a buyer would, for the minimum contract term, what happens on cancellation, data on exit, uptime and support response times, time to go live, SSO, and how you compare with the two rivals you lose to most. Write down every one that turns up nothing.
2. Give each one an owner outside marketing. Finance owns contract terms, legal owns data on exit, customer success owns onboarding time, security owns SSO. Ask each owner for three sentences a buyer could read.
3. Publish a range, not a promise. A typical range plus what moves it beats silence every time. If you decide to keep something behind sales, say so plainly. That turns a dead end into a gate, which is a choice you made.
4. Put the answer where the buyer is standing. Contract and pricing questions belong on or near the pricing page (here is how we think about pricing on your website). Security answers belong on the security page, not in a deck. An answer buried deep in the docs does less work than one on the page the buyer is reading, which is the point of answers hidden in the docs.
5. Reuse what presales already wrote. Some of these answers may already exist in security questionnaires and RFP responses. Our security questionnaire playbook for presales covers how to keep a library of them.

Then keep listening. Your 20 questions are a starting point; real buyers will ask ones you didn't predict, and most buyer questions arrive when your sales team isn't around.

This is where I should say plainly that I work on a product built for this. RepX, Storylane's website agent, answers buyer questions on your site around the clock, works out whether the visitor is a fit, and hands sales-ready buyers to your team. Qualified leads and conversation summaries flow to HubSpot, Salesforce and Slack, so the questions buyers actually ask become a running list your team can work through. It suits demand gen and web teams who want their site to answer the ownerless questions instead of leaving them blank.

The one idea to keep: most website silence isn't strategy, it's a question nobody owns. On Monday, ask your own site the ten questions finance, legal and customer success get asked, and give every dead end a name and an owner.

If you want to see how RepX handles the questions your pages don't answer yet, book a demo with our team.