How to compare two growth proposals
Proposals are hard to compare when each one argues differently. Put both into the same seven rows, then compare cost per unit of expected result and how quickly each one can be proven wrong.
ComparisonExample
| Element | Proposal A: new landing page | Proposal B: search restructure |
|---|---|---|
| Evidence cited | Landing page converts 1.1% vs 2.4% site average | 38% of spend on queries with no sales |
| Diagnosis | Visitors don't see pricing | Broad match pulls informational searches |
| Proposed action | Rebuild page with pricing above the fold | Exact and phrase match; add 120 negatives |
| Budget and owner | $2,400 · agency | $600 setup + current spend · freelancer |
| Expected range | +0.3 to +1.0 pts conversion | −15% to −30% cost per customer |
| Stop and scale | Stop if below 1.2% after 2,000 visits | Stop if customers fall >20% in 21 days |
| Measurement | A/B split, same traffic | Before/after, 21 days each, same budget |
How to choose
- Prefer the proposal whose evidence you can verify in your own data.
- Prefer the cheaper way to be proven wrong: B costs less and reports in 21 days.
- Reject any proposal without a stop condition. It has no defined next decision.
Use this with your own numbers. Keep the evidence private, attach a budget and a review date.
Build my opportunity brief