Start with the decision, not the dashboard
The first measurement question is what the business needs to decide. More visitors may be useful for an awareness objective, but a service business seeking sales conversations needs to know whether enquiries are relevant. Choose the outcome before choosing the report.
Write down what counts as success and who checks it. A completed lead form is a measurable action. A qualified lead is a later business judgment that needs explicit criteria, such as service fit, a valid need and a reachable contact. Do not treat the two as synonyms.
Choose and verify important events
Google Analytics describes a key event as an event measuring an action particularly important to business success. For lead generation, a successful form submission can be one such action. Test the successful submission separately from a form opening or a button click; otherwise the count can describe intent rather than completion.
Document the event's trigger and check it with a real test submission. Confirm that duplicate page loads or repeated clicks do not inflate the count. Keep personally identifiable form details out of analytics events; the CRM is the place to manage contact records under an appropriate privacy process.
Connect the enquiry to its next stage
An enquiry is the start of follow-up, not proof of revenue. Record whether the lead is qualified, whether a conversation takes place and whether the opportunity progresses. A campaign that generates fewer enquiries can still be useful if those enquiries fit the business better.
For a simple review, compare campaign spend, completed enquiries and qualified leads over the same period. Cost per qualified lead is spend divided by the number of qualified leads; when that number is zero, report that no qualified leads were recorded rather than presenting a misleading cost figure. Agree on the qualification criteria before comparing channels.
Explain uncertainty in the result
Attribution reports describe recorded touchpoints within their measurement settings; they are not a complete account of every influence on a sale. Consent choices, offline conversations and the length of a buying decision can affect what the reports contain.
When reviewing a change, note the date range, traffic source, measurement settings and changes to the offer or landing page. Separate an observed association from a claim that one campaign caused the outcome. This makes a report more useful than a large click total without context.