Map the real buying route
Identify where people discover the offer, seek information, contact the business and complete a purchase. Calls, visits, partner referrals and events may form part of the path. Agree which decisions require measurement before building a complicated attribution model.
Collect useful information proportionately
Use consistent enquiry categories, appropriate campaign references or other practical signals where suitable. Ask staff to record context without creating excessive friction for the customer. Explain data handling and obtain specialist advice where needed; this guide does not prescribe a legal basis for tracking.
Reconcile results with their limits
Compare enquiries and sales using consistent periods and definitions. Report unmatched records, missing information and possible double counting. Combine operational evidence with customer feedback to decide where to improve the experience. An incomplete but honest view is more useful than a precise-looking attribution claim built on assumptions.
Put the decision into practice
- Map online and offline decisions.
- Choose a few useful signals.
- Train the receiving team.
- Reconcile periods and definitions.
- Report missing matches and uncertainty.
Bring the business question, the evidence already available and the constraints that matter. Discuss your next step with Nancy Chou.
Map the customer record across touchpoints
A customer may research online, ask a question by phone and purchase offline. Describe which events are observable and which cannot be reliably linked. Avoid summing platform-attributed revenue with store or sales totals without reconciliation. Use privacy-appropriate identifiers and access practices rather than collect unnecessary personal information.
Distinguish journey description from causal credit
A recorded interaction can explain part of a journey without proving that it caused the purchase. Define the operational decision the reporting must support: follow-up, information repair or budget testing. Where identity matching is incomplete, report the limitation and use an appropriate sample or bounded experiment. Do not turn an incomplete cross-channel record into a falsely precise return calculation. Pipeline attribution: credit, coverage and causality.
An enquiry that becomes an offline purchase
Hypothetical decision exercise. Suppose a customer visits a campaign page, telephones for details and later buys through a sales colleague. The records may show several interactions but only one transaction. Reconcile what is observable and define the follow-up decision the report should support. Do not award the full sale to each channel or claim that the recorded sequence proves incremental impact.