Product Learning Lab
Concept

Funnel Analysis

Reading a journey step by step to find where users drop and why.

Aggregate numbers tell you the product is leaking. A funnel tells you where.

A journey broken into steps

A funnel is a user journey split into ordered stages — visit, sign up, activate, purchase — with the count of users who reach each one. The shape it makes is the point: every step is narrower than the last, and the sharpest narrowings are where the product is losing people.

The value isn't the chart, it's the question it forces. A conversion rate of 4% is a fact you can't act on. A funnel showing that 70% of signups never complete onboarding is a fact you can.

Core insight

A funnel converts a vague outcome ("conversion is low") into a located problem ("the drop is between step 2 and step 3"). Location is what makes it fixable.

The step tells you where; it doesn't tell you why

The trap is stopping at the biggest drop and calling it the answer. A cliff between "added to cart" and "purchased" could be surprise shipping cost, a broken payment step, or a buyer who was only ever comparison-shopping. The funnel narrows the search; qualitative work and segmentation close it.

Segmenting the same funnel — by source, device, new vs returning — often reveals the drop lives in one slice, not the whole. A checkout that's fine on desktop and collapses on mobile reads as a mild average until you split it.

Takeaway

Use the funnel to find the biggest, earliest drop — then treat that step as a question to investigate, not a conclusion to act on.

Keep going
Skill · Analytical

Funnel thinking

The skill of applying this in practice.

Skill · Analytical

Metric drop diagnosis

When a funnel step suddenly worsens.

Concept

Activation

The step a funnel most often exposes.