How to Read a Funnel

One thing I learned pretty early in product management is that a funnel looks simple until you actually have to figure out what is happening inside it.

You have users entering at the top, users moving through different steps, and eventually some of them reaching the outcome you care about. Seems straightforward, right? But when conversion suddenly drops at one stage, the real work starts, because the funnel tells you where something is happening, but not necessarily why.

And that distinction is important.

Start With the Journey, Not the Numbers

Before looking at conversion percentages, understand what each step in the funnel actually represents.

For example, imagine a B2B product funnel:

Sign Up → Create Project → Invite Users → Launch Project → Complete First Workflow

Each step represents a meaningful user action.

If you immediately notice that only 40% of users create a project, you might conclude that project creation is a problem.

But first ask what the user is actually trying to accomplish at that stage.

Maybe they don’t understand what to do next. Maybe they need information from someone else. Maybe the product is asking for too much information before they can continue.

The numbers are telling you where to investigate, not what the solution should be.

Look at Conversion Between Each Step

One of the easiest mistakes is looking only at the final conversion rate.

Suppose 10,000 users enter your funnel and 1,000 eventually complete the desired action. Your overall conversion is 10%.

That number is useful, but it doesn’t tell you much about where users are struggling.

Instead, break it down:

• 10,000 start
• 7,500 complete Step 1
• 5,000 complete Step 2
• 2,500 complete Step 3
• 1,000 complete Step 4

Now the picture becomes much clearer.

The biggest drop isn’t necessarily where the largest number of users disappear in absolute terms, but where the relative conversion rate changes significantly.

A drop from 10,000 to 7,500 is 25%.

A drop from 2,500 to 1,000 is 60%.

That second step deserves much more attention.

Don’t Assume Every Drop Is a Product Problem

This is where product managers can get into trouble.

You see a 50% drop between two steps and immediately start thinking about redesigning the screen, changing the button, or simplifying the workflow.

But users may be dropping for completely different reasons.

Maybe the next step requires an approval from their manager. Maybe they don’t have the information required. Maybe they’re not actually expected to complete that step.

I’ve seen funnels where a large drop looked like a UX problem, but after talking to users it turned out to be an intentional behaviour.

So before fixing the funnel, understand why the funnel exists in that shape.

Segment the Funnel

An overall funnel can hide some very interesting behaviour.

Break it down by meaningful segments such as:

• New vs returning users
• Customer size
• User role
• Geography
• Acquisition channel
• Device
• Product plan
• Use case

You might discover that your overall conversion is 30%, but enterprise customers convert at 60% while smaller customers convert at 15%.

Now you have a very different product question.

Instead of asking, “How do we improve conversion?”

You can ask, “Why does this particular segment behave differently?”

That is a much better starting point.

Look at Time Between Steps

Funnels aren’t always about whether users complete a step. Sometimes how long it takes them matters just as much.

If most users move from Step 1 to Step 2 within a few minutes, but some users take three days, there may be friction that isn’t visible through conversion alone.

This becomes particularly important in B2B products, where users often depend on other people to complete parts of a workflow.

A funnel with 80% conversion but seven days between steps can be a bigger problem than a funnel with 70% conversion that users complete in ten minutes.

Combine the Funnel With Qualitative Data

A funnel tells you what happened.

It doesn’t tell you what the user was thinking.

That’s why I rarely look at funnel data in isolation when investigating a meaningful drop. I want to combine it with session recordings, support tickets, customer interviews, surveys, and sometimes even conversations with sales or customer success teams.

The strongest insights usually appear when the quantitative and qualitative data point toward the same problem.

Final Thought

Reading a funnel isn’t about finding the biggest percentage drop and immediately building a solution.

It is about understanding the journey, identifying where behaviour changes, looking for meaningful segments, and then investigating what could be causing that change.

A funnel is essentially a map of user behaviour.

But a map can tell you where people stopped walking. It can’t tell you whether they got lost, reached their destination, changed their mind, or simply had somewhere else to go.

The job of a Product Manager is to find that part out.