Google AdsA campaign can produce inexpensive signups and very few customers your sales team can serve. Before changing copy or bids, check whether the event being counted represents the outcome the business actually wants.
This guide prepares a measurement specification for a fictional B2B scheduling product. It does not connect a CRM, upload personal data or change a bidding goal. Use the lead-stage measurement worksheet to agree on definitions with the sales and measurement owners first.
Define each stage with evidence
Google distinguishes qualified leads and converted leads as conversion categories that can represent later steps in a lead process. Your business still has to define what those steps mean and how the evidence reaches the measurement system.
For this example, a form submission creates an enquiry. A qualified lead means a reviewer has confirmed that the company fits the product's supported use case and has an actual scheduling need. A converted lead means the agreed sales milestone has occurred. Those definitions are illustrative, not a universal SaaS funnel.
| Stage | Evidence required in this example | Owner |
|---|---|---|
| Enquiry | A completed, non-test submission | Marketing operations |
| Qualified lead | Confirmed use case and recorded qualification decision | Sales operations |
| Converted lead | The chosen closed-sale milestone | Revenue operations |
Avoid definitions that depend on an undocumented judgement. “Good lead” cannot be reproduced by another reviewer. Write the specific criteria, the event timestamp to use and what happens when a decision is reversed.
Show how the chosen metric changes the conclusion
Consider this fictional comparison over a mature reporting window. The same spend produced different volumes and quality of enquiries. No causal or statistical conclusion is implied by these small example counts.
| Measure | Campaign A | Campaign B |
|---|---|---|
| Ad spend | $1,000 | $1,000 |
| Enquiries | 100 | 50 |
| Qualified leads | 10 | 20 |
| Cost per enquiry | $10 | $20 |
| Cost per qualified lead | $100 | $50 |
Campaign A looks cheaper when judged by enquiries. Campaign B looks cheaper when judged by qualified leads. The calculation changes because the business event changes. Use the CPC and CPA calculator with the actual clicks and relevant event count, then write the event name beside the result.
This still does not establish profitability. Sales conversion, contract value, delivery cost and the time needed to collect revenue remain separate questions. Do not assign an invented lifetime value to every qualified lead to make the dashboard look more commercial.
Specify the handoff from business records to reporting
Document the supported conversion import method, identifier requirements, event name, timestamp, timezone, value treatment and owner. Consult Google's enhanced conversions for leads configuration guide for that particular option and its prerequisites. Select an implementation that matches the actual consent and data setup.
Use synthetic records to test the process before sending real customer data. Check duplicate handling, rejected rows, delayed qualification and later corrections. A successful upload does not by itself prove that every eligible event matched, that the business definition is correct or that the reporting window is mature.
Keep personal information out of campaign URLs and public worksheets. A stable internal record ID can support reconciliation in an appropriately controlled system; it should not become a public UTM parameter.
Review goal settings before changing optimization
Google's primary and secondary conversion action guide explains how those settings interact with reporting and bidding, including the special treatment of custom goals. Review the campaign's actual selected goals rather than assuming that changing a label affects every campaign in the same way.
Before an owner approves a switch, confirm that the downstream event is reliably collected and arrives within a useful time frame. Record which campaigns will change, the previous configuration and the intended review date. Do not combine an unverified import with a broad goal change and then use the following day's performance as proof that it worked.
Reconcile a mature cohort and keep exceptions visible
Choose a cohort old enough for the normal sales process to develop. Compare the eligible CRM records, accepted imports and reported outcomes, documenting why the totals can differ. Separate unresolved measurement gaps from genuine sales disqualifications.
If most rejected leads came from an unsupported use case, review the search terms and negative keyword workflow. If the right audience signed up but failed to activate, examine the product journey before blaming the ad. The point is to locate the next question accurately.
Bring the agreed definition, validation evidence and proposed changes into the Google Ads preflight or a supported AdRails review. A shared definition gives the campaign owner something concrete to approve and monitor.
Sources checked October 14, 2026. The funnel and monetary amounts are illustrative planning examples.