Marketing funnel: connect acquisition with product use
A marketing funnel shows how people move from discovering a product to taking a meaningful action. Connecting it with activation and retention helps a product team see whether acquisition brings customers who can use the product and want to return.
What a marketing funnel represents
A funnel is a simplified sequence of actions taken by a defined group. It can describe a short path from landing page to signup, or a longer one through first use, payment and renewal. The scope depends on the question you want to answer.
People do not always follow that sequence. They may return on another device, enter through a referral or buy before trying the product. Use separate paths where the journey differs materially, and do not interpret every missing step as a failed customer experience.
Translate funnel stages into observable actions
Top, middle and bottom of funnel (TOFU, MOFU and BOFU) are useful shorthand for awareness, consideration and conversion. They are broad categories, not event definitions. A product team needs to specify what a person actually does at each stage.
For a trial product, that might mean visiting an offer, creating an account, completing a useful task and starting a paid plan. Retention then measures whether the intended behavior recurs within a suitable window. If purchase comes before first use, reflect that order instead.
- Discovery: a relevant person encounters the offer. Reach and visits describe exposure, not product value.
- Consideration: the person examines fit, for example by reviewing use cases or trying the product.
- Activation: the person completes a defined action that indicates initial value, with the relationship to later use still to be tested.
- Payment: a verified customer or account meets the chosen paid conversion definition.
- Retention: the acquired group repeats the meaningful behavior or renews over a specified period.
Set the denominator, sequence and time window
A step conversion rate divides people who reach the next step by eligible people at the preceding step. Keep the unit consistent: users, accounts and sessions produce different answers. State whether the steps must happen in order and how long the transition may take.
For a cohort-based conversion measure, give each cohort the full observation window. Comparing users who signed up yesterday with a group that has had a month to pay creates an artificial decline.
An analytics report may include users who enter at any stage or require them to start at the first step. GA4 calls these open and closed funnels. Choose the setting that matches the question rather than comparing rates from different definitions.
Worked example: from visits to useful product use
In a hypothetical trial product, a measured group contains 10,000 visitors, 1,000 trial starters, 400 people who complete the core task and 100 who become paying customers. The trial-to-task rate is 40%, task-to-paid is 25% and visitor-to-paid is 1%.
Suppose 60 of those paying customers perform the core task in week four. That is 60% week-four retention among the 100 paying customers, not among all trial starters. The group and return event belong beside the reported rate.
If task completion rose from 400 to 500 and task-to-paid stayed at 25%, the scenario would produce 125 paying customers. The unchanged downstream rate is an assumption to test: helping a different mix of trial users complete a task may change who pays and stays.
Investigate why people stop
A drop-off tells you where to look, not what caused it. Before proposing a redesign or a new campaign, check whether the event is missing, the step is optional or the user has not had enough time to continue.
Then combine journey data with observation and customer conversations. Someone may leave setup because they cannot import their data, because the offer does not fit their task, or because they need approval from a colleague. These causes require different responses.
Prioritize a problem by the number of affected eligible users, the importance of the blocked task and the confidence in the explanation. The largest percentage drop is not automatically the most valuable product problem.
How AARRR adds product questions
AARRR groups growth questions under acquisition, activation, retention, referral and revenue. It helps teams give product use and repeat use their own attention instead of ending the discussion at signup.
These categories do not form a universal chronological funnel. A user can pay before activation or invite a teammate during setup. Treat them as questions to answer, then build event sequences around the actual product journey.
Check the measurement before changing the product
In GA4 funnel exploration, step conditions, sequence and time limits control who is counted. Test the report with known journeys, including a user who skips an optional step and one who takes longer to continue. A valid purchase missing from one funnel can indicate a definition mismatch.
Where possible, reconcile paid conversions with billing or CRM records. Account creation, device switching, consent and incomplete instrumentation can interrupt the analytics history. Record the limitation instead of assuming that every unmatched event represents abandonment.
Use a customer journey map to add the context that event counts cannot show: what people are trying to do, what information they need and what interrupts the task.
Choose an experiment and follow its downstream effect
State the affected audience, the observed obstacle and the change expected to help. Choose a primary outcome and the later outcomes that must not deteriorate. Shorter signup may increase registrations while leaving activation or paid conversion unchanged.
For example, test whether a sample dataset helps eligible trial users complete their first planning task. Measure that task completion, then compare payment and repeat use for cohorts old enough to observe them. A favorable first-step result is not yet evidence of healthier growth.
| Stage | Observed action | Definition to record |
|---|---|---|
| Visit | View the relevant offer | Eligible audience and counting unit |
| Trial | Create a trial account | New accounts only and duplicate policy |
| Activation | Complete the core task | Successful completion event and allowed window |
| Payment | Start the paid plan | Billing confirmation and customer unit |
| Retention | Repeat the core task | Return event, cohort and observation period |
FAQ
How many stages should a marketing funnel have?
Use enough to locate the decisions or tasks you can act on. Broad awareness, consideration and conversion categories are a starting vocabulary. Add product-specific steps when they answer a distinct question.
What is a good conversion rate between stages?
There is no rate that applies to every audience and event definition. Compare like-for-like cohorts, investigate changes in their behavior and consider the economics of the resulting customers.
Is a marketing funnel the same as a sales funnel?
A sales funnel typically follows prospects through the sales process. A marketing funnel can begin earlier, while a product funnel focuses on actions inside the product. Their boundaries should match the team's question and data.
Does funnel drop-off prove a usability problem?
No. Missing events, optional paths, low intent and delayed decisions can also cause it. Inspect the measurement and observe the task before choosing a product fix.
Can a signup improvement make growth worse?
It can attract or admit a different mix of users, increase service cost or leave fewer people reaching the intended value. Follow activation, payment and retention as well as the initial signup rate.
Sources
Discuss your product's next step
I help define product priorities, plan launches and investigate where users drop out.
