Give each stage its own evidence
To track whether a blog produces customers, connect the article visit to a confirmed booking and then the sales outcome. Keep reading, clicking a booking button, completing a booking, attending a call and paying as separate stages. Each answers a different question.
Search Console shows search visibility and clicks. Analytics shows observed website activity. Your booking system confirms the appointment; your sales records establish attendance, qualification, proposals and wins. One dashboard cannot replace missing records from another part of that process.
Start with a small event and record map you can verify. More events do not necessarily provide better attribution. A contact who discovers an article, watches a video and returns a week later may leave an incomplete trail. Preserve that uncertainty instead of inventing a perfect single-source journey.
Define the path you want to observe
Use stable labels and an agreed definition of “qualified.” If an appointment is rescheduled, preserve its relationship to the same opportunity. If a customer books twice, your sales report needs a rule for identifying that duplication.
| Stage | Evidence | Avoid counting as |
|---|---|---|
| Article discovery | Observed landing page and acquisition source | A qualified enquiry |
| Video or offer route | Click to the relevant offer; observed video event if implemented | A completed booking |
| Confirmed appointment | Booking-system confirmation and unique appointment ID | A held sales call |
| Held and qualified | Sales owner records attendance and fit | A won customer |
| Won and paid | Recorded deal and payment outcome | Assumed revenue from a page view |
Preserve the path between sites
When the blog, video page and booking flow use different domains, review how the source is preserved. Google’s cross-domain measurement guidance describes using the same web-stream tag and verifying the linker where applicable. Subdomain cookie settings and self-referrals also need checking.
Use campaign parameters on externally distributed links to identify the campaign. Keep internal navigation from overwriting the original acquisition source; store article or route context as separate fields. A link from your own blog to your own offer is a route through the funnel, not a new external acquisition campaign.
Booking forms and webhooks may support source or metadata fields. Verify the fields your actual provider supports and the values that reach the confirmed booking. Do not send names, email addresses or other personal details inside analytics URLs or event labels.
Verify a booking event against the booking record
A button click can be measured immediately, while a completed booking needs evidence that the booking system accepted it. Where available, use the booking confirmation callback or verified webhook and a deduplication identifier. Retrying a webhook should not create another booking outcome.
Test a successful booking, abandoned flow, failed booking and reschedule in the provider’s supported test environment. Confirm the expected analytics event and sales record for each case. Seeing a booking interface load is not evidence that the completion event works.
Ask the sales owner to record attended, no-show, cancelled and rescheduled outcomes. If that information is not maintained, the report can tell you about bookings but not attended calls. Label the limitation plainly.
An example that keeps attribution honest
Illustrative example: 200 observed organic article sessions lead to ten offer-route visits, three confirmed bookings, two attended calls and one paid customer. Report all five counts with their measurement window. Do not call the ten route visits ten leads or infer payment from the confirmed booking.
If the paid customer later says they first heard about you on LinkedIn, retain that answer alongside the observed article landing. It may describe a different stage of discovery. Count one customer, with two pieces of attribution evidence—not one organic customer plus one LinkedIn customer.
Small counts can move sharply. One more attended call changes a two-call result by 50%, but that does not establish a better topic or funnel. Look at the absolute counts, lead quality and the time allowed for the buying cycle.
Review content by the decisions it helps buyers make
Separate new and old articles, brand and non-brand search, and markets with different offers. Compare pages over similar published ages. A recently indexed article and an established page have had different opportunities to be discovered.
Keep unattributed customers and unobserved visits visible. Consent choices, device changes and later direct visits can leave gaps. A “How did you hear about us?” answer can help, but it is self-reported evidence rather than a recovered click log. Search queries in GSC are aggregated; they are not an identity-level map of which query produced a customer.
Use the worksheet below to agree the event and record owners. Rainlight’s pipeline review starts with those boundaries so your content report can show what happened between reading an article and becoming a customer.