Define the result before reading a dashboard
Agree what makes an enquiry relevant: the requested service, actual coverage, a genuine contact route and any qualification rules the business uses. Keep spam, staff tests and duplicate submissions identifiable in restricted business records.
Search clicks, visits, phone/email activations, accepted submissions, messages received and qualified enquiries are different measurements. Search Console’s clicks and impressions describe search activity; they do not establish a sale or even a contact attempt.
Illustrative example: a supplier’s company-name searches can lead to product enquiries about that supplier, while a service provider may be seeking commissioning enquiries. More traffic to a portfolio page is not automatically more qualified demand for the provider.
Test the contact path before commissioning new content
In a safe test environment, submit a clearly labelled valid request and inspect the server response, validation, confirmation, email/provider result and intended inbox. Repeat an invalid submission and a simulated sending failure. Confirm staff know where requests arrive.
Application validation and email dispatch are separate operations; Django’s email tools support explicit send-error handling and test backends. A successful-looking screen is not enough evidence of actual receipt. On the live system, use an approved test route and owner-authorized recipient rather than sending unexpected customer-style messages.
Check phone numbers and email destinations on mobile. A working link still does not prove someone answered the phone or received an email.
Locate the first stage with missing evidence
- Little relevant search exposure: check indexing/access, intended query groups and whether the preferred service page answers that need.
- Exposure but few clicks: inspect the actual result context, query mix, title/description accuracy and the page shown. Do not infer a defective snippet from CTR alone.
- Visits but little contact activity: review offer clarity, actual coverage, relevant proof and whether the next action can be found and used.
- Contact activity but few received requests: investigate validation, sending, destination accounts and staff handling before assuming an SEO problem.
- Received requests but poor fit: clarify service boundaries and qualification, and review which pages/channels attract that audience.
Use comparable date ranges and the same country, device and channel definitions. With very few events, inspect longer trends and individual safely redacted workflow evidence instead of treating a small percentage movement as a stable trend.
Check that an event means what the report says
GA4’s automatic form interactions describe starting/submitting a form. Inspect whether a custom accepted-enquiry event instead depends on server acceptance, and whether automatic and custom handlers are being added together. Refreshing a confirmation page or loading tracking twice should not manufacture another accepted request.
Verify a safe test in the intended property with DebugView or appropriate collection evidence. A public measurement identifier or a local queued event does not prove ingestion. Existing consent choices, blocked transport, short visits and configuration can affect collection; absence from Analytics is not proof of zero enquiries.
Keep personal information out of Analytics, including event fields, page URLs and titles. Business qualification belongs in restricted enquiry handling; report aggregate outcomes without publishing messages, names or contact details.
Fix the page decision supported by the evidence
Ask whether a suitable visitor can identify the service, confirm coverage, understand key exclusions and find a sensible next step. Use genuine relevant work; keep illustrations and demos distinguishable from client delivery. Explain what information a quote request needs and how it will be handled.
Review the actual mobile path, including menus, overlays, form errors and confirmation. If the right service content already exists, repair the broken step rather than producing more city pages or repeating keywords. The service-website checklist helps review the necessary information.
Record changes and compare like with like
Keep a dated record of the affected URLs, observed failure, implemented change and verification result. Recheck the contact workflow after a release. Compare an equivalent post-release period with the preceding period using the same filters, and note any concurrent campaigns, seasonal changes or tracking changes.
Review accepted and qualified enquiries alongside visibility and interactions, not as interchangeable totals. A repaired form can resolve a confirmed defect without proving it caused a later ranking or enquiry change. For help, discuss the evidence and measurement scope or a specific website repair.


Comments
Leave a comment
Comments are reviewed before publishing.