When ad platforms, analytics tools, and a CRM report different conversion or revenue totals, this reconciliation checklist helps you isolate the cause, document the agreed source of truth, and avoid changing campaigns based on misleading data.
Overview
Ad attribution discrepancies are common because each system may answer a slightly different question. An advertising platform may report conversions credited to an ad interaction, an analytics tool may organize sessions and journeys under its own attribution model, and a CRM may count leads only after they meet a status or revenue condition. These numbers can all be internally correct while still failing to match.
The goal of reconciliation is not to force every system to show the same total. The goal is to understand why the totals differ, decide which number is appropriate for each business decision, and identify genuine tracking failures. A useful reconciliation process compares the same date range, event definition, time zone, currency, attribution window, and reporting status before anyone investigates campaign performance.
Start with a simple variance record for each comparison:
- System and report: for example, an ad platform conversion report, analytics purchase event, or CRM qualified-lead report.
- Metric definition: what counts as a conversion, qualified lead, sale, or revenue event.
- Reporting period: the date range and time zone used by the report.
- Attribution settings: click-through, view-through, data-driven, last-click, or another selected method.
- Observed variance: the absolute and percentage difference.
- Confirmed cause and owner: the explanation, next action, and person responsible for follow-up.
For broader campaign quality checks, pair this process with a PPC campaign audit checklist. Tracking discrepancies should be reviewed alongside targeting, landing-page, budget, and conversion setup issues.
Checklist by scenario
When ad platform conversions exceed analytics conversions
- Confirm whether the ad platform counts conversions after an ad click while analytics counts only sessions or events that successfully loaded.
- Compare the conversion action definitions. One system may count form starts, phone calls, imported offline conversions, or engaged visits while the other counts completed purchases.
- Check whether view-through conversions are included in the ad platform report. If analytics requires a recorded site visit, the totals will not be directly comparable.
- Review consent and privacy signals. Some users may not be eligible for analytics measurement, while another system may receive a modeled or imported conversion.
- Look for duplicate conversion actions, such as both a browser event and an imported CRM event being included in the same platform total.
- Check reporting delays, especially for offline conversions, lead-status changes, and revenue updates.
When analytics conversions exceed ad platform conversions
- Verify that the analytics event is assigned to paid traffic rather than organic, direct, referral, email, or unattributed traffic.
- Check URL parameters and redirects to ensure campaign information is preserved from the ad click to the landing page.
- Review the ad platform attribution window. A conversion may occur outside its selected click or view window even though analytics records the event.
- Check whether the platform imports only selected conversion actions while analytics records every event with the same name.
- Compare conversion counting rules. A platform may count one conversion per ad interaction, while analytics may count multiple completed events from one user or session.
- Confirm that the platform account, campaign, and conversion action are connected to the same website or measurement property.
When CRM leads or revenue do not match either system
- Define the CRM milestone being reported: submitted lead, accepted lead, qualified opportunity, closed deal, or collected revenue.
- Check whether duplicate leads are merged, rejected, or excluded in the CRM but retained in the advertising or analytics report.
- Confirm that offline conversion imports use a stable identifier and the correct event date.
- Review lead-status timing. A lead created during one reporting period may become qualified or closed in a later period.
- Reconcile refunds, cancellations, taxes, discounts, and currency conversions before comparing revenue.
- Check whether sales entered manually are linked to the original campaign, click, or user record.
When only one campaign, channel, or market shows a variance
- Compare the affected campaign's tags, landing pages, forms, phone numbers, and checkout flow with a campaign that reconciles normally.
- Check account-level and campaign-level conversion settings rather than assuming they inherit the same configuration.
- Review differences in device, geography, language, consent behavior, and landing-page templates.
- Use search-term and campaign change records to identify whether the discrepancy began after a URL, form, bid, or tracking change.
For a repeatable cross-platform view, use a cross-platform ad reporting template and preserve each system's native figures alongside normalized comparison fields.
What to double-check
Attribution windows and models
Write down the selected click-through and view-through windows for every platform being compared. Then record whether the report uses a last-touch, position-based, data-driven, or other model. A conversion credited to one channel in one system may be assigned to another channel elsewhere. Do not label a model difference as a tracking error until the settings are aligned or the difference is intentionally documented.
Time zones and conversion dates
Use the same time zone and date boundaries wherever possible. Also note whether a system reports the conversion on the interaction date or the conversion date. This distinction can create apparent daily discrepancies even when totals become closer over a longer period.
Click, view, and assisted interactions
Separate click-through conversions from view-through conversions and assisted conversions. They represent different relationships between advertising exposure and the eventual outcome. A combined total should never be compared with a report that includes clicks only.
Event duplication and counting rules
Test whether a confirmation page reload, repeated button click, browser retry, or server-side retry can send the same event more than once. Compare event IDs, transaction IDs, order numbers, and lead IDs where available. Also record whether each platform counts every conversion or only one conversion per ad interaction.
Consent, identity, and missing parameters
Check whether consent choices affect cookies, advertising identifiers, analytics storage, or CRM matching. Review whether redirects, link shorteners, checkout domains, and cross-domain journeys preserve campaign parameters. A practical UTM tracking governance process can reduce preventable attribution loss by standardizing names and testing links before launch.
Currency, revenue, and import timing
Confirm the currency displayed in every report and the currency used in the underlying event. Decide whether revenue is gross, net, discounted, refunded, or tax-inclusive. Then note when offline revenue enters each system. Revenue reconciliation is incomplete if the event counts match but the value definitions do not.
Common mistakes
- Comparing unlike metrics. “Conversions” is not a sufficient definition. Name the event and status behind the number.
- Changing tracking before preserving evidence. Export the affected reports, settings, event samples, and timestamps before making fixes.
- Using one platform as the universal truth. Use the ad platform for delivery and optimization context, analytics for onsite behavior, and the CRM for lead quality or realized revenue when those roles fit the business process.
- Ignoring lag. Keep a provisional reporting status for recent periods instead of treating incomplete data as final.
- Fixing symptoms with campaign changes. Do not pause keywords, alter bids, or reallocate budget solely because a conversion total looks unusual. Validate measurement first. The PPC budget reallocation framework is more useful when its inputs have been reconciled.
- Failing to record the resolution. Add the confirmed cause, affected dates, system owner, and prevention step to a shared change log. This prevents the same investigation from restarting at the next monthly close.
When to revisit
Run a lightweight reconciliation before each monthly reporting close and a deeper review before seasonal planning cycles. Revisit the checklist whenever a form, checkout, CRM field, consent mechanism, analytics property, tag, landing-page template, or offline import changes. It is also appropriate after an unexplained conversion swing, a new advertising channel launch, a domain migration, or the introduction of a new reporting dashboard.
At the start of each review, choose a stable comparison period and a recent test period. For each metric, record the platform totals, analytics totals, CRM totals, variance percentage, and status: aligned, expected difference, under investigation, or requires correction. Set an internal tolerance that reflects the maturity and importance of the measurement system rather than treating every small difference as an incident.
Finish with three actions: preserve the evidence, assign the next owner, and state whether campaign decisions should wait. If the cause is confirmed, update the conversion definition or tracking documentation, test a known conversion from click to CRM, and add the result to the reconciliation log. When the workflow or tools change again, repeat the same comparison instead of relying on memory.