Why do marketing reports show different numbers?
Meta, Google Analytics, Google Ads and your sales system can report different results because they measure different actions, apply different attribution rules and may record outcomes in different reporting periods. Multiple platforms can also claim credit for the same sale.
A discrepancy does not automatically mean the campaign failed or the report is wrong. However, it should not be dismissed as “normal” without checking the definitions and implementation.
For Singapore SMEs, the goal is to explain the differences and make sound budget decisions, not force every dashboard to display the same total.
What is marketing attribution?
Marketing attribution is the process of assigning credit to the interactions that precede an enquiry or purchase. An attribution model determines how that credit is allocated; an attribution window determines how far back an eligible interaction can be considered.
Google Analytics lets businesses select settings governing how key-event credit is assigned, including the reporting model and lookback window. Meta’s standard attribution setting can consider eligible ad impressions, clicks or engagements.
These reports offer different views of the customer journey. They are not interchangeable with a record of completed transactions.
| Reporting source | Useful question it answers | Important limitation |
|---|---|---|
| Meta or Google Ads | Which eligible outcomes receive credit under this platform’s rules? | Credit can overlap with another platform’s report |
| Google Analytics | How do measured website interactions and channels relate to key events? | Results depend on implementation and report settings |
| CRM or lead log | Which enquiries became qualified opportunities or customers? | Source information may be incomplete |
| Order or sales system | Which transactions were recorded and what was their value? | Transactions alone do not explain marketing influence |
1. Your customer may interact with several channels
Imagine someone discovers a skincare product through an Instagram ad, later searches the brand on Google, and finally places an order.
Depending on the applicable settings and observed interactions, both advertising platforms may credit that purchase. Your store still records one order.
This is why adding platform-reported sales together can overstate the number of unique purchases. The same issue can affect reported revenue.
Use unique order identifiers to reconcile transaction totals where available. Treat each platform’s attributed results as its own perspective, rather than an exclusive allocation of all sales.
2. Attribution windows answer different questions
A short window asks whether a conversion followed a recent interaction. A longer window allows more time between the interaction and the outcome.
That distinction matters when comparing an impulse purchase with a considered service enquiry. Customers may need more time to evaluate a renovation package than to buy a familiar household item.
Record the settings used in each report, including whether impressions or engagements can receive credit. Check your account’s available options rather than assuming a universal default.
Changing a window can change reported performance without changing the number of actual customers. Document reporting changes so they are not mistaken for business growth.
3. A conversion is not always a sale
One dashboard might count form submissions while another counts purchases. Your CRM may report only qualified enquiries.
Before comparing results, agree on an event dictionary:
- Enquiry: a successfully submitted request.
- Qualified lead: an enquiry meeting agreed suitability criteria.
- Customer: a lead that has completed the defined purchase milestone.
- Revenue: a clearly specified amount, such as paid order value or recognised sales.
For an e-commerce brand, distinguish placed orders, paid orders, cancellations and refunds. For a service business, distinguish bookings from attended appointments and paid engagements.
Keep event names and definitions consistent. A low cost per enquiry cannot be compared directly with a cost per paying customer.
4. Date ranges and revenue definitions may differ
A customer can click in September, enquire in October and pay in November. Reports based on interaction dates, event dates or payment dates can place that journey in different periods.
For monthly reviews, record the date basis and align time zones where possible. Keep reporting cut-off times visible and allow for processing delays.
Also confirm whether revenue includes discounts, delivery charges, taxes and refunds. A gross checkout value and net completed sales value are different metrics, even when they relate to the same order.
Do not rename both “revenue” and assume the comparison is fair.
5. Some gaps come from measurement problems
Not every difference is an attribution issue. Check for duplicate purchase events, missing tags, broken campaign parameters or a checkout journey that loses source information.
Google documents several reasons advertising clicks and Analytics sessions differ, including multiple clicks within one session and visits that end before the page loads. A click, a session and a purchase are different units.
Test the measurement journey, including mobile forms and payment redirects. Check whether one confirmed transaction generates the intended event once and whether its identifier, currency and value are correct.
Avoid changing attribution rules to conceal an implementation error.
6. Create a reconciliation report, not a summed dashboard
Start with deduplicated business records for actual orders or customers, while checking that those records are complete. Then show channel-attributed results separately.
Explain known differences in a short reconciliation note: overlapping credit, reporting dates, event definitions, missing identifiers or data still processing.
For leads, carry campaign information into your lead log where feasible and follow outcomes through to qualification and purchase. For orders, compare tracked purchase identifiers with actual order records.
Attribution indicates credited contribution. It does not, by itself, establish how many sales would have disappeared without the advertising. Controlled experiments can help assess that separate question when feasible.
A practical first step for your next report
Select one completed reporting period. Align definitions, dates and settings, then inspect a manageable sample of orders or leads.
Separate explained differences from unresolved tracking gaps. Fix the largest measurement problem before making major budget changes.
Reliable reporting does not mean identical numbers everywhere. It means understanding what each number represents and how it supports a business decision.
Unsure which marketing numbers to trust? Contact Brands Win Agency to discuss campaign measurement and reporting. Growth System 90 brings performance ads, influencer and KOL marketing, and content strategy together under one business goal.
FAQs
Why does Meta report more sales than my sales system?
Possible causes include overlapping channel credit, different reporting periods, duplicate events or different transaction definitions. Check the configuration and transaction records before deciding which explanation applies.
Can I add Meta and Google Ads conversions together?
Not as a count of unique customers or orders unless overlap has been resolved. Both platforms may credit the same conversion under their own rules.
Is GA4 always more accurate than advertising platforms?
No. GA4 provides a different measurement view, not an automatically superior one. Review the event implementation, attribution settings and report scope.
Which report should I use for actual sales?
Use deduplicated order, payment or sales records with a clearly defined status and period. Validate those records and use marketing reports to understand credited channel contribution.
Does attributed ROAS prove that my ads caused the sales?
No. Attributed ROAS reflects revenue credited under the chosen measurement rules. Establishing additional sales caused by advertising requires a separate evaluation, such as an appropriate controlled test.

