Where to start
- Define an enquiry before comparing totals
- Keep observation separate from attribution
- Record unknown sources honestly
- Agree on one operational next action
Agree on the unit your team will count
Write down what counts as a new enquiry, a suitable enquiry and a follow-up. Decide how duplicate contacts, spam, existing customers and wrong-service requests are handled. Use the same definitions in the reporting period and comparison period. A changed definition can change the total without any change in demand.
Give the intake team ownership of classification and the marketing team ownership of the website measurement explanation. The shared report needs aggregated counts and a description of the rules, not contact names, message contents or confidential customer histories.
Keep three evidence columns separate
Search Console shows search performance, such as clicks and impressions. Website analytics can show recorded sessions and configured events subject to the implementation and available data. Your operational system can confirm that an enquiry was actually received. These totals need not match, and none should be silently substituted for another.
Attribution assigns credit under a model and its available touchpoints; it is not proof that one channel caused a sale. Record the reporting window, source definition and known gaps beside any attributed figure. If the source is unknown, keep it unknown rather than allocating it to the largest traffic channel.
Copy this monthly handoff worksheet
Use one row per priority service or landing-page group. Do not join anonymous search queries to named individuals. Keep any record-level work inside the authorized business system and share only the minimum aggregated result.
- Period and comparison: dates, timezone and comparable service scope.
- Search evidence: observed clicks/impressions and source date, or unavailable.
- Website evidence: the exact event definition and whether receipt is verified.
- Operational evidence: received enquiries, classification rules and duplicate handling.
- Attribution: known source category, method and unassigned remainder.
- Explanation: tracking change, outage, campaign, seasonality or service-capacity note.
- Action: one owner, one change and one review date.
Reconcile a hypothetical mismatch
Imagine a hypothetical report shows 24 contact-button clicks but the team finds 10 new enquiries in its approved records. This is not automatically 14 lost leads. Some visitors may click more than once, decide not to contact, use a different route or return later. First confirm what the event records and how duplicates are handled.
If the contact path is broken, repair and verify it. If receipt is working but classification is inconsistent, fix the reporting handoff. Do not manufacture test customers to make the totals line up, and do not report this example as a result for TorontoSEO.agency or a client.
End the report with a decision, not a claim
Write the strongest conclusion supported by the evidence and its limitation. For example: the team cannot yet compare suitable enquiries by landing page because source capture is incomplete. The next action is to validate the existing measurement design and agree on the classification process, not to announce an SEO decline.
Keep the prior worksheet so the following review can distinguish completed work from repeated recommendations. If you want help with the measurement scope, bring your public website and a summary of the definitions to a consultation; keep private customer data out of an initial marketing enquiry.