A website report is useful when it helps you decide what to do next. For an enquiry-led business, start with visits, contact actions and enquiries your team actually received. Keep sales in a separate record. A button click alone cannot tell you whether somebody sent a message, qualified for your service or paid.
This guide is for an existing business website that brings people to a form, phone number, email address or WhatsApp conversation. It is a measurement brief for the owner and the person maintaining the site, rather than a promise that installing analytics will generate business.
Define the result before choosing a dashboard
Write down what counts as a genuine enquiry: for example, a new person asking about a service you actually provide. Decide how to handle spam, supplier pitches, repeat messages and existing customers. Keep those categories separate so a busy inbox does not become an inflated new-business figure.
Then agree who reviews the inbox and records the outcome. The website can record a contact action; your team usually needs to confirm what happened after it. A useful record can be as simple as date, enquiry category, source if known, current stage and eventual outcome. Keep customer details in your controlled business system.
Match each question to the evidence
| Owner's question | Evidence to review | What it does not prove |
|---|---|---|
| Are people finding us in Google Search? | Search Console impressions, clicks, queries and landing pages | That a visitor became a customer |
| What happens after people arrive? | Analytics visits and selected page or interaction events | That every visitor is measured |
| Are people trying to contact us? | Tested form or contact-link events | That the message arrived or a call connected |
| Are suitable enquiries arriving? | Form storage, inbox or call records reviewed by the team | That every enquiry came from the website |
| Did the business win work? | Agreed order, payment and business records | That the website alone caused the sale |
Google explains the different roles of Search Console and Analytics. Their figures need not match: a search click and a measured visit are different observations. Compare dates, time zones, page scope and definitions before investigating a discrepancy. Do not force the totals to agree.
For WhatsApp, label a measured link action as a contact click. A visitor may open the app and leave without sending anything. For a form, distinguish a browser interaction from a submission saved by the form service. If enquiries are missing, use the contact form troubleshooting guide.
Prove that a count means what its label says
Ask the maintainer for an event list in plain language: the name, the exact trigger and the question it answers. Google Analytics lets you mark important actions as key events; the label does not validate the underlying trigger.
Agree a controlled test with the person who receives enquiries. On desktop and mobile, check the chosen route from the real landing page. Record the expected action, analytics evidence and the receiving system's result. Include an abandoned form and a failed submission. Neither should appear as a confirmed enquiry simply because a button was pressed.
Google's enhanced measurement documentation describes automatic interaction events, including outbound clicks and forms. Check the options and behaviour on your actual site. Do not assume every contact route is covered, or add another tag before checking whether it would duplicate an existing count. Keep names, email addresses and message contents out of analytics events and page URLs.
If your own team's visits distort a small sample, review internal traffic filters. Test before activation: excluded data is permanently unavailable. Any tracking setup also needs the consent and privacy choices appropriate to the business; this guide does not determine those requirements.
Review one comparable period at a time
Choose a review rhythm that fits the volume. A monthly review is an editorial starting point, not a required reporting interval. Use the same definitions and distinguish full periods from partial ones. Note campaigns, closures, site changes and tracking outages.
Start with enquiries received and whether they fit your business. Then work backwards: which landing pages and contact routes were involved, and where is evidence missing? If traffic increased but enquiries did not, inspect the audience and contact path before buying more traffic. If contact clicks increased but received messages did not, first test that route.
Illustrative arithmetic only: 400 measured visits and 12 contact clicks gives a 3% contact-click rate when both counts cover the same defined period and scope. If your team received five suitable enquiries, that remains a separate count. Without a reliable link between records, do not attribute all five to those 12 clicks or describe 3% as a sales conversion rate. These figures are not an industry benchmark.
Ask for a small, usable handover
Ask for owner-controlled access to the relevant properties, the event definitions, a dated test record, known measurement gaps and the person responsible for repairs. Add a short report that ends with one decision and the evidence needed to judge it next time.
You do not need a wall of charts to begin. You need to know what was counted, what reached the business and what remains unknown. Google's lead measurement tutorial is a useful implementation reference; the business still needs to reconcile the results with its actual enquiries and sales.
