Likes are not leads: measure what your website actually records

The report shows visits and interactions, but nobody can identify which requests reached support or became qualified opportunities. Likes may indicate interest in a post; they do not establish a commercial enquiry. Separate exposure, action, saved requests and conversation stages. Each measure answers a different question and needs a real source.

Define a valid enquiry

A button click may not lead to submission. A request can be retried. A saved lead may be spam or unsuitable for the service. Agree what counts as a valid request and who classifies its later stage before building the report.

Search Console and analytics observe different parts of the journey. Search clicks are not an exact substitute for website visits, and different periods or filters do not describe the same population.

Verify the record you already have

Check whether the form saves origin, selected service and limited campaign fields. Test submission, retry and inbox visibility. In a hypothetical retry, the same idempotency key should recover one request; counting both attempts as leads would inflate reporting.

Use stages appropriate to the actual workflow, with responsibilities for contact, qualification and proposals. Current status alone does not reconstruct time spent in each stage. Agree history and criteria separately. Opening an email link does not mean the opportunity closed.

Preserve context without persistent visitor identification

Names, phone numbers and messages do not belong in URLs or analytics events. Limited attribution can accompany a consented contact request without a persistent visitor identifier. An article supplied in the journey is declared context, not proof it caused an engagement.

I can structure reporting through dashboards and BI and integrations, using the existing commercial records. New tracking requires authorisation and a defined purpose. If analytics is disconnected, say so rather than display zero as a measured visit total.

Start with a question such as how many valid requests received follow-up in the period. Then inspect recorded services and origins. This provides useful operational evidence without inventing financial returns from likes.