A B2B website with few enquiries does not automatically have a traffic problem. A working form may not be measured, an educational article may attract visitors with no buying intent, or a suitable customer may not understand within ten seconds who the service is for. Diagnose in a fixed order: measurement, traffic, message, proof and action.
Buying more traffic before those five layers work mainly makes the leak larger.
Diagnostic map
| Signal | Likely layer | First check |
|---|---|---|
| Sales receives enquiries analytics does not show | Measurement | Complete form, phone and calendar events yourself |
| High blog traffic, few service-page visits | Traffic intent | Compare landing pages and onward clicks by topic |
| Service page is visited but quickly abandoned | Message | Test audience, problem and outcome above the fold |
| Cases are viewed but contact stays low | Offer or risk | Check next step, price expectations and objection handling |
| Form is started but not completed | Friction | Test fields, errors, privacy copy and mobile use |
Layer 1: are you measuring the right enquiries?
Start with a real test on mobile and desktop. Submit every form, click phone and email links, book an appointment, and verify that confirmation, internal notification and CRM registration all arrive.
Only mark actions as key events when they represent business value. Analytics can show key events per landing page, but it cannot know by itself whether a submitted form became a qualified enquiry. Where possible, feed later sales status back into the measurement: relevant, meeting, proposal, won or rejected.
Watch for duplicate events, internal employees, spam and cookie consent. A dashboard showing twenty “conversions” may represent five real enquiries.
Layer 2: is the right traffic arriving?
Split organic traffic into at least three intents:
- learn: the visitor wants an explanation or definition;
- compare: the visitor is researching approaches, prices or suppliers;
- decide: the visitor wants a quotation, specialist or concrete solution.
An article can receive thousands of educational visits without producing a direct enquiry. That does not make it worthless: it can build expertise and point to the next step. Do not expect buying behaviour from every search query.
For each landing page, identify the service that logically follows. An article on spreadsheet risk can point to when a web application becomes the better option, not straight to a generic contact form.
Layer 3: does the right customer understand the offer?
A hero saying “digital solutions for tomorrow” communicates almost nothing. Answer three questions at the top:
- Which type of organisation is this for?
- Which concrete problem does it solve?
- What can the visitor do next?
Use customers' words, not just internal terminology. “API orchestration” may be technically correct; “stop retyping webshop orders into Exact” makes the outcome visible.
Show the page to an outsider for five seconds, then ask who the service is for and what it delivers. If their answer differs from yours, the message is not sharp enough.
Layer 4: is there enough proof for a risky decision?
B2B purchases often have long cycles and several decision-makers. A prospect must believe not only that you can deliver, but that the process is manageable.
Strong proof includes:
- a case with starting point, approach, timescale and measured outcome;
- a screenshot, architecture diagram or concrete deliverable;
- names and relevant experience of the delivery team;
- a clear process, ownership terms and aftercare;
- boundaries: when the solution is not a fit.
A row of logos without context or an anonymous “great collaboration” quote removes little risk. One developed case can be more convincing than ten generic testimonials.
Layer 5: does the next step fit the buying stage?
“Contact us” makes visitors invent what happens next. Make the step specific:
- have a workflow analysed in 30 minutes;
- receive a scope sheet for a website proposal;
- discuss a technical integration with an architect;
- have an existing proposal or migration plan reviewed.
Explain what the visitor should prepare, who joins, how long it takes and what they receive afterwards. Reduce uncertainty without artificial scarcity.
Check form friction
Ask only for information needed to plan the next step. A budget or timeline question may be useful, but explain why. Test labels and errors with keyboard and screen reader, a narrow mobile screen, network failure, double-clicking, data retention after an error, and the clarity of privacy and confirmation copy.
A four-week improvement test
- Week 1: repair measurement and record a baseline per landing page.
- Week 2: rewrite audience, problem and outcome on one important service page.
- Week 3: add one strong proof block and a specific next step.
- Week 4: analyse qualified enquiries, not just event counts.
Keep acquisition channels and major campaigns as stable as possible during the test. Otherwise you will not know which change caused the effect.
The goal is not more forms but better conversations
A shorter form can increase submissions while also attracting more irrelevant leads. Judge the whole chain: suitable visitor → clear service → credible proof → appropriate next step → qualified sales conversation.
For a new build, also use the website proposal scorecard.