The decision this guide supports: which limiting condition deserves scarce team capacity now? A long idea backlog is not a roadmap. Good prioritization begins with a diagnosis of the business mechanism currently restricting valuable growth.

Use the economic scorecard from G02 to locate the greatest value at risk or the strongest obstacle to progress. Seek both internal and customer evidence. Do not automatically pick the highest percentage drop, since a small-volume step can look dramatic while a more modest issue affects most profitable orders.

Diagnose before scoring opportunities

Examine the nine growth levers as connected explanations. Poor checkout conversion may reflect unsuitable acquisition traffic, misunderstood offers, fulfillment concerns or an interface defect. Flat revenue may reflect weak demand, unavailable inventory, pricing trade-offs or unmeasured cancellations. State the main hypothesis and at least one serious rival explanation.

Validate measurement quality at each step. A broken purchase event, mismatched time zones or incomplete ad cost can make a problem appear larger or smaller than it is. When evidence is blocked, a bounded validation task may outrank a speculative conversion project.

Make a defensible opportunity estimate

Estimate the number of eligible customers or transactions affected, the plausible commercial effect per unit and the degree of confidence in the assumptions. Separate upper-bound opportunity size from expected benefit and from what can be credibly measured. Risk and effort are not mere inconvenience; a payment change may carry security or operational consequences requiring additional QA.

Use a transparent priority method. ICE, RICE, PIE or a customized weighted matrix can support comparison if scales and inputs are declared. Do not combine raw scores from incompatible matrices as if the number were universal. Allow a named decision owner to override recommendations with a written reason.

Build a three-part 90-day roadmap

Days 1–30: stabilize the evidence, validate the leading constraint and deliver one reversible improvement or experiment where justified. Days 31–60: execute the strongest approved intervention, monitor guardrails and address implementation dependencies. Days 61–90: evaluate results, refine the operating process and prepare the next decision based on the outcome.

This is a planning pattern, not a promise that every significant change fits into ninety days. Platform migrations, inventory cycles and low-traffic experiments may require a different horizon. Keep the first cycle small enough to learn rather than filling the calendar with parallel initiatives.

Worked scenario: retention versus new acquisition

Consider a fictional subscription retailer with respectable new-customer volume but weak second-order conversion. A growth team proposes a paid-media expansion because new orders are easy to measure. Cohort data suggests that increasing the second purchase rate of existing customers may offer more contribution at lower marginal marketing cost.

The team validates the cohort definitions, studies onboarding and post-purchase feedback, then compares a retention experiment with the acquisition expansion. The roadmap focuses on the better-supported constraint and keeps the alternative visible for later review.

Roadmap acceptance checklist

• One named commercial objective and non-negotiable guardrail

• A primary constraint supported by traceable evidence and alternatives

• A prioritized shortlist with estimates, confidence and capacity

• Approved work with owners, QA, dependencies and dates

• A review routine that may stop, revise or reopen the plan

Decision rule: when a high-priority initiative cannot be explained in relation to the current constraint, remove it from the active roadmap. G04 continues the diagnosis by showing how market research and positioning can reveal whether the real problem is the customer promise rather than the interface.

← Previous guide · Next guide →