Conversion and analytics: what happens after the click.

Usually several causes at once: a goal firing on a page load rather than a genuine submission, two tags counting the same event, consent refusals removing a share of visitors entirely, cross-domain tracking breaking at a third-party form, and phone or email enquiries never being counted at all. We reconcile both sides and fix the measurement before drawing any conclusion from it.

The benchmark is unhelpful and often misleading, because it depends entirely on traffic mix and on what you count as a conversion. A site drawing specific, high-intent search traffic may convert at several per cent; one with broad awareness traffic will be far lower and may be performing better commercially. The useful comparison is your own rate over time, segmented by channel, measured on qualified enquiries.

Sometimes, and we will tell you when the answer is no. Reliable testing needs enough conversions to reach significance in a reasonable period, which many B2B sites do not have. Where testing is not viable we use qualitative evidence — session review, form field analysis, user testing and sales feedback — and make reasoned changes rather than pretending an underpowered test proved something.

Significantly, and most businesses have never quantified it. Refused consent means those visitors are not measured, so your reported traffic and conversions are a subset of reality. We implement consent compliantly, measure the acceptance rate so you know the size of the gap, and use consent mode so the loss is modelled rather than invisible.

The enquiry form, almost always — specifically its length, the fields that are required, and whether it asks for information a buyer is unwilling to give at that stage. After that, clarity about who you serve, and making phone contact genuinely easy for buyers who prefer it. These are usually larger wins than redesigning a page.

With call tracking: a dynamically inserted number that ties the call back to the session and channel that produced it, passed into the CRM alongside web enquiries. Without it, every phone enquiry appears as direct traffic, which systematically understates the channels that drive high-intent visitors.

Enough conversions to reach significance in a reasonable period, which in practice means a few hundred conversions a month rather than a few hundred visitors. Below that we use qualitative evidence — session review, form field analysis, user testing and sales feedback — and make reasoned changes rather than pretending an underpowered test proved something.

Almost certainly, and usually downwards at first, because the old numbers were counting things twice or counting page loads as conversions. That is uncomfortable and it is the point: you cannot optimise against figures that do not match what sales received. We document what changed so the discontinuity is explainable.

Yes, if the identifiers are carried through. Source and campaign are captured at first touch, stored against the CRM record and still present when the deal closes, so you can report revenue by original source rather than by last click. For long B2B cycles that distinction changes which channels look worthwhile.

The enquiry form, almost always — its length, the required fields, and whether it asks for information a buyer will not give at that stage. After that, clarity about who you serve, and making phone contact genuinely easy for buyers who prefer it. Both beat redesigning a page.