# SaaS funnel analysis worksheet

Copy into your preferred editor. Define units, progression and cohort maturity before calculating. A scenario is a model with assumptions, not a forecast or proof of achievable lift.

## Measurement definition

- Decision this analysis should inform:
- Unit: user / account / another defined unit:
- Eligible starting population:
- Cohort entry date/range:
- Observation window and minimum cohort age:
- Required order of events:
- Identity matching approach and limitations:
- Excluded internal/test records:
- Data sources and known definition changes:
- Validation checks performed:

## Funnel

| Step | Exact event / condition | Eligible count | Conversion from previous step | Quality / data concerns |
|---|---|---|---|---|
| Start | | | — | |
| Step 2 | | | | |
| Step 3 | | | | |
| End | | | | |

Step conversion = next-step count / eligible previous-step count. Use rates as decimal fractions in multiplication: 5% = 0.05.

## Scenario comparison

| Candidate change | Baseline rate | Scenario rate | Expected downstream outcome | Assumptions / evidence | Effort / risk |
|---|---|---|---|---|---|
| | | | | | |

Relative lift = (new rate - baseline rate) / baseline rate. Percentage-point change = new percentage - baseline percentage. Relative lift is undefined when baseline is zero; investigate counts and measurement instead of dividing by zero.

## Completed fictional example

Ordered mature cohort with resolved user identities: 2,000 visitors → 100 signups → 40 activated → 10 paid. The rates are 5%, 40%, and 25%; overall conversion is 0.5%.

| Scenario | Calculation | Expected paid customers |
|---|---|---:|
| Baseline | 2,000 × 0.05 × 0.40 × 0.25 | 10 |
| Signup rate reaches 6% | 2,000 × 0.06 × 0.40 × 0.25 | 12 |
| Activation reaches 50% | 2,000 × 0.05 × 0.50 × 0.25 | 12.5 |
| Activated-to-paid reaches 35% | 2,000 × 0.05 × 0.40 × 0.35 | 14 |

Fractional outcomes are model expectations. Real customer counts are whole numbers. These scenarios hold other rates and traffic constant; changes can affect audience quality and downstream rates. Figures are not benchmarks.

## Investigation and intervention brief

- Where the experience appears to fail:
- Evidence from data QA:
- Evidence from successful and unsuccessful users:
- Proposed explanation and alternatives:
- Change to investigate or make:
- Expected outcome and assumptions:
- Quality/reliability guardrails:
- Owner, dependencies and effort:
- Review method and window:
- Result, confidence and next decision:

## Final checks

- Same unit and comparable maturity across steps?
- Counts reflect ordered progression rather than unrelated monthly totals?
- Missing, duplicate and internal events checked?
- Impact compared with effort, evidence and risk?
- Small-sample or before/after evidence described honestly?


---
Original resource by Jon Bishop. CC BY 4.0. Illustrative examples are not customer results.