Conversion Tracking
Part of Paid social reporting and account reviews
Checking whether reported efficiency comes from existing customers
Check first and repeat orders beside paid social results, while keeping customer mix separate from unverified ad attribution.
A low reported cost per purchase does not show the cost to acquire a new customer. Check the customer status of confirmed orders in business records, then which campaigns received platform credit. Without a suitable link between those records, customer mix warns or reassures about the business total, but does not prove which ads produced the orders.
Define the order population
Choose a period and define a confirmed order. Decide how payment, cancellations, refunds, guest checkout, merged accounts and multiple stores affect the count and customer identity. For each eligible order, classify the customer as first-time, returning or unknown under the business's rule.
Before relying on a customer report for distinct confirmed orders in a period, check its definitions.
Count first and repeat orders
Use distinct eligible order IDs to count and value each status group. Put orders with unreliable identity history in a separate row. Show the counts behind any percentage.
| Business population | Record | Limit |
|---|---|---|
| First-time customer orders | Confirmed first orders under the chosen identity rule | Does not establish that advertising caused them |
| Returning-customer orders | Confirmed later orders under the same rule | Does not show whether those customers would have bought anyway |
| Unknown status | Eligible orders without dependable history | Cannot be assigned confidently to either group |
This is an order-population view. Do not label the first row “paid social acquisitions” merely because it falls in the same reporting period as the ads.
Check the advertising context
Separate spend and attributed results for acquisition, remarketing and existing-customer re-engagement. Inspect audience sources and exclusions that applied during the period. A prospecting label does not prove every recipient was new.
Audience exclusions may not cover customers absent from their source or unmatched by the platform.
Where a permitted, technically reliable order-level reconciliation is available, identify reported events that connect to known orders. State the method, coverage and unmatched share before describing the matched subset's customer mix. Some platform conversions may be modelled or otherwise unavailable for an order-level match.
Do not move personal details into a reporting file merely to make the chart complete. Handle personal information in line with applicable privacy obligations.
State what the check supports
If attributed efficiency looks strong but confirmed first orders are scarce, investigate campaign mix, exclusions and changes in attribution settings. The result does not establish that existing customers generated a particular share of attributed purchases unless the records support that connection. Even a matched first order does not show that the ad caused an additional sale. If that causal question governs a substantial spend decision, use a suitable comparison design.
Steps to Verify Customer Origin of Reported Efficiency
- Define confirmed order eligibility (e.g., payment complete, no refund)
- Classify customersfirst-time, returning, unknown based on identity rules
- Match platform-reported events to known orders using reliable reconciliation
- Separate acquisition, remarketing and re-engagement campaign results
- Assess attribution settings and audience exclusions for gaps
