Agree on a qualification rule
More visits can coexist with fewer suitable enquiries. Before comparing pages or channels, define what makes an enquiry useful to the business. For a service agency, a working rule might require a requested service the team actually provides, a project within its delivery market and enough information to assess the request. These are illustrative operational criteria, not a universal scoring system.
Write the rule, its effective date and the reviewer responsible for applying it. Include an undecided status for requests that need clarification. Do not turn an unanswered enquiry into a rejection merely to complete a spreadsheet, or let different reviewers quietly apply different standards to competing channels.
Keep acquisition and qualification as separate stages
Google’s recommended event reference distinguishes generate_lead, for a generated lead, from qualify_lead, for one marked as meeting qualification criteria. The event names do not supply your criteria or automatically connect your sales records to Analytics. Any integration needs its own documented implementation and verification.
For an initial review, use an authorized internal record with one row per enquiry and a stable local reference. Record arrival date, review date, status and the evidence for the status. Keep contact details in the operational system; use aggregate totals in the SEO report rather than copying private enquiry text into analytics fields.
Compare cohorts with equal time to mature
Choose a cohort by enquiry arrival date and allow the same review interval for each period. A recent week may appear worse simply because its requests have not been reviewed yet. Show received, qualified, not qualified and pending counts together, and explain which denominator each rate uses.
Record source or landing-page information only where it is supported by the available evidence. Leave unmatched records visible. Do not assign all unknown enquiries to organic search to make totals reconcile. Keep the traffic table alongside the enquiry table without pretending that incomplete matching is a complete visitor-level journey.
Illustrative comparison with pending cases
Consider two fictional cohorts that have both had fourteen days for review. Cohort A contains 40 enquiries: 12 qualified, 20 not qualified and 8 pending. Cohort B contains 30: 15 qualified, 10 not qualified and 5 pending. Qualified enquiries as a share of all received are 30% and 50%, respectively.
If the team reports the rate among reviewed enquiries instead, the denominators become 32 and 25, giving 37.5% and 60%. Neither definition should be substituted silently. B has more qualified enquiries in this example despite fewer enquiries overall, but the figures do not establish statistical certainty, sales revenue or a causal effect of SEO.
Use the difference to choose an investigation
Read rejection reasons in aggregate to identify a specific page question: does the offer attract work outside the service scope, or does the page omit information prospects need? Review a small authorized sample with the responsible team before changing copy. A count alone cannot tell you which explanation is correct.
Keep the qualification definition stable during the comparison and document any revision as a break in the series. Finish with the page to inspect, the evidence supporting that choice and a follow-up date. Use the resulting brief for a focused SEO review, while continuing to report visits as reach and qualified enquiries as a separate business-stage measure.