Interactive funnel analysis

Funnel Growth Lab

A worked example of finding where applicants drop off, estimating the value of one focused change, and setting clear checks before launch.

TL;DR

Compare customer groups, find the biggest drop-off, test a realistic improvement, and connect added activations to an editable revenue estimate.

All applicants

  1. Started10,000
  2. Eligible7,600
  3. Offer viewed5,900
  4. Verification started4,100
  5. Verification complete2,850
  6. Activated2,420

Sample monthly group · 24.2% activation

01 · Find the problem

Find the biggest drop-off before choosing a fix.

Switch customer groups to see where people leave. Conversion and drop-off come directly from the sample counts.

90-second test

Check the finding.

Change the inputs and see whether your choice still holds.

  1. 01Switch customer groups and see where the largest percentage drop-off moves.
  2. 02Change the improvement and revenue assumptions, then check the estimates.
  3. 03Review the test idea and safety checks before choosing what to do.
Customer group

The total of the three sample customer groups.

Selected group10,000applications started
Start-to-finish conversion24.2%started → activated
Biggest drop-off69.5%Offer viewedVerification started

People at each step

  1. Started10,000
    100.0% of selected group
  2. Eligible7,600
    76.0% from prior stage
  3. Offer viewed5,900
    77.6% from prior stage
  4. Verification started4,100
    69.5% from prior stage
  5. Verification complete2,850
    69.5% from prior stage
  6. Activated2,420
    84.9% from prior stage

Step-by-step rates

Conversion and drop-off for each step in the All applicants group
StepConversionDrop-offApplicants lost
StartedEligible76.0%24.0%2,400
EligibleOffer viewed77.6%22.4%1,700
Offer viewedVerification started69.5%30.5%1,800
Verification startedVerification complete69.5%30.5%1,250
Verification completeActivated84.9%15.1%430

Biggest drop-offs

Start with the largest leak

The list ranks each step by drop-off rate and keeps the number of people lost in view.

  1. 1

    Offer viewedVerification started

    1,800 applicants lost

    30.5% drop-off
  2. 2

    Verification startedVerification complete

    1,250 applicants lost

    30.5% drop-off
  3. 3

    StartedEligible

    2,400 applicants lost

    24.0% drop-off
  4. 4

    EligibleOffer viewed

    1,700 applicants lost

    22.4% drop-off
  5. 5

    Verification completeActivated

    430 applicants lost

    15.1% drop-off

02 · Estimate the value

Test a realistic improvement, not a promised result.

Change one step and one revenue assumption. Later-step behavior stays the same so the math is easy to check.

+5 pp

Percentage points change the rate directly. A move from 40% to 45% is a five-point gain.

$100

Change this input to connect conversion to money. It is an assumption, not measured customer value.

Estimated additional activations

+174per sample monthly group
Estimated step conversion
74.5%
More people moving ahead
295
Later-step completion held at
59.0%
Illustrative revenue estimate
$17,400

174 × $100 = $17,400. Excludes repeat use, refunds, costs, and profit.

See the math

Improvement used: +5.0 percentage points

5,900 × 0.050 × (2,420 ÷ 4,100) = 174

People entering the step × the point increase × the share who finish later steps = estimated added activations.

  • Assumes the group size and mix stay the same.
  • Assumes added applicants finish later steps at the current rate.
  • Does not include capacity limits, season changes, or test uncertainty.

03 · Decide what to test

Turn the finding into a test that can prove you wrong.

The funnel shows where people leave, not why. Customer research must support the idea before launch.

Test idea

Make the hardest all applicants step easier

30.5% drop-off · 1,800 applicants lost

What may be happening
Applicants put off verification because they do not know why it matters, how long it takes, or which documents they need.
Smallest useful test
Test an upfront verification checklist with a time estimate and an explanation of why each item is needed.
Primary metric
Offer viewed → Verification started conversion
Decision rule
Keep the change only if more people start without more support questions or quick cancellations.
What this test will not do
Do not remove required verification or alter the offer itself.

04 · Protect customers and the business

A conversion gain only matters if quality stays strong.

The main metric shows possible growth. Safety checks show whether the change is worth keeping.

Safety checks

Proposed test safety metrics
MetricThresholdWhy it matters
Verification qualityNo meaningful dropGrowth cannot come from weaker required checks.
Support contact rateNo more than a 5% increaseA faster funnel should not create more confusion for support.
Seven-day cancellation rateNo more than +1 percentage pointMore activations only help if people still want the product.
Largest customer-group conversion gapMust not widenOverall growth should not hide a worse result for one group.

Checks before launch

  1. 01

    Research

    Is this the real problem?

    Customer sessions confirm why people stop, not only where the numbers fall.

  2. 02

    Design

    Can people understand it?

    Applicants can explain the requirement, next step, and result during testing.

  3. 03

    Engineering + Data

    Can we trust the tracking?

    Views, steps, errors, and customer-group events match before the test begins.

  4. 04

    Legal / Compliance

    Does it follow the rules?

    The copy and flow keep all required notices, consent, eligibility, and checks.

  5. 05

    Operations

    Can support handle problems?

    Support steps, owners, escalation paths, and rollback rules are clear.

What comes next

Data shows where to look. Customer research shows whether the problem is real.

Before the roadmap changes, validate this analysis through customer sessions, a prototype test, and a measured experiment.