Measure SEO as a sequence: relevant visibility, visits to useful pages and qualified enquiries. A change in a chart is a prompt to investigate. It is not proof that the website has acquired customers.
This guide provides a copyable report and a decision table for a small service business. Keep the same page group and reporting settings between reviews so that the comparison remains useful.
Define a genuine enquiry
Agree what qualifies before reading the numbers. For example, a service business might count a person requesting a quotation for a service it provides in an area it covers. Record known owner tests, spam and duplicates separately.
A WhatsApp click shows an attempt to open a contact route. It does not prove that a message was sent. A saved form submission shows that the enquiry reached storage; it does not prove that the person is a suitable customer or that an email notification arrived. Verify the steps that your business relies on.
Use each source for the question it answers
| Source | Useful question | Limitation to record |
|---|---|---|
| Search Console | Which Google searches and pages received impressions or clicks? | Clicks are not unique customers; some query data is withheld |
| Website analytics | What did measured visitors do after arrival? | Consent, blocking and configuration affect what is recorded |
| Enquiry register | Which messages were received and which were relevant? | Someone must classify tests, duplicates and lead quality |
| Sales record | Which qualified enquiries became agreed work? | A sale may happen later or through a different contact route |
Search Console and Analytics do not measure the same event, so their totals do not have to match. Do not force one report to reproduce the other.
Set up a UK service-page comparison
For a UK acquisition review, define the reporting group before collecting totals:
- In Search Console, select Web, United Kingdom and a complete date range. Record the exact settings and whether the newest data is preliminary.
- List the service pages and supporting guides being assessed. Keep pages for different products or audiences outside this group. Retain a whole-site overview separately.
- Record whether the figures describe the property or individual pages. Do not sum page-row impressions and present that number as the property total: they use different counting rules.
- Record the rule used to separate brand searches. Query filters exclude anonymised searches, so filtered non-brand rows are not a complete account of all discovery.
- In Analytics, inspect the corresponding landing pages and country with the intended session source and medium. Google organic Search is a different slice from all organic search engines.
The Search Console data guide explains aggregation and preliminary data; its query guidance explains omitted searches. Match settings where possible and record the remaining differences rather than expecting the tools to agree exactly.
Copy this 28-day report
Compare equal-length periods using the same search type, country or device filters and page group. Record seasonal events and incomplete data. Use a longer view when the sample is too small to support a decision.
| Field | Current period | Previous comparable period |
|---|---|---|
| Start and end dates | Record dates | Record dates |
| Search type and filters | Record settings | Use the same settings |
| Included pages or service group | List the URLs | Use the same group |
| Brand-search rule | Record how company-name searches are separated | Use the same rule |
| Relevant search impressions | Record value | Record value |
| Search clicks | Record value | Record value |
| Click-through rate | Record value | Record value |
| Average position | Record value and query mix | Record value and query mix |
| Measured organic sessions | Record value and consent limitations | Record value and limitations |
| Received qualified enquiries | Confirm against the register | Confirm against the register |
| Known tests and duplicates | Record separately | Record separately |
| Website changes and external factors | List dated changes | List dated changes |
CTR is clicks divided by impressions. For illustration, 20 clicks from 1,000 impressions gives 2% CTR. Those figures are an arithmetic example, not Nolixia results or a target for every website.
Read combinations of signals
| Observed pattern | Investigate next | What it does not establish |
|---|---|---|
| An intended page is absent from the index | Inspect its status, access, discovery and content purpose | A new title alone will solve indexing |
| Relevant impressions rise but positions remain weak | Compare intent, useful detail, evidence and relevant links | More impressions automatically mean more customers |
| Useful positions but few clicks | Inspect the actual query, result type, title and page relevance | The meta description is the sole cause |
| Measured visits rise but received enquiries do not | Test the offer details, phone layout and contact journey | All traffic is suitable or tracking is complete |
| Enquiries arrive but few qualify | Review service area, scope, price expectations and lead classification | Search traffic is necessarily the only problem |
Average position can change because the query mix changes. A site appearing for more new queries may show a different average even when an established query improves. Review the important service queries and pages individually.
Keep lead stages outside the traffic totals
Use one internal request reference to follow a received enquiry. A small register can record the landing page or source when known, requested service, received status, qualification decision, proposal and agreed-work status. Leave a stage unknown or pending when there is no evidence. Do not guess which search query produced an individual lead.
For example, a fictional record could read: request DEMO-01; service-page source known; message stored; email receipt not yet checked; suitability awaiting review; no proposal or agreed work. That is a received record with unresolved checks, not a verified qualified lead or customer.
Google's recommended lead events distinguish submission, qualification and becoming a customer; those events need implementation and do not verify themselves. Keep names, email addresses, telephone numbers and message text out of ordinary Analytics event parameters and URLs, following Google's personal-information guidance. Store the operational details in the authorised business system.
If visits are recorded but enquiries are absent, follow the service-business enquiry diagnosis before treating the issue as a need for more traffic.
Keep a dated change log
Record the URL, what changed, why it changed and what was verified. Separate implementation checks from business outcomes: a working form is verified by delivery; more qualified enquiries require later evidence.
Do not label an owner test as a conversion from a customer. Exclude or label internal tests consistently, and avoid treating a day or two after a release as a completed SEO evaluation. Record both the report's date range and its last available data.
Use the SEO audit checklist when the numbers reveal a specific issue. For a launch, start with the new website checklist.
Sources
- Search Console Performance report
- Using Analytics and Search Console together
- Google guidance on requesting a recrawl
- Search Console aggregation and preliminary data
- Search Console query and page groupings
- Google Analytics recommended lead events
- Avoiding personal information in Analytics
For a commercial page receiving search appearances but few visits, use the impressions-but-no-clicks diagnosis to work through query fit, position and result presentation before choosing a focused change.
