Starting From a Single Growth Constraint
Every good growth plan starts by naming the single biggest constraint on growth right now — not a list of five equally weighted priorities. Is the constraint top-of-funnel volume, a leaky activation step, weak retention, or an inefficient paid channel? Everything else in the plan should trace back to addressing that one constraint, because a plan that tries to fix everything at once usually fixes nothing well.
Translating Modules Into a Coherent Plan
Each earlier module in this course — channels, CRO, attribution, retention, budget — is a lens, not a separate to-do list. A strong capstone plan pulls the relevant lens for your specific constraint: if retention is the constraint, lean on the lifecycle and cohort material; if it's acquisition efficiency, lean on the channel and budget material. Don't force every module into the plan just because you learned it.
Setting a 90-Day Horizon With Real Milestones
Ninety days is long enough to see a real result and short enough to force prioritization. Break it into three roughly month-long phases: build and instrument, test and iterate, and scale what worked. Each phase needs its own concrete milestone — not just "make progress," but a specific, checkable deliverable.
Defining Success Metrics You Can't Argue With Later
A metric like "improve growth" isn't a success metric — it can't be checked. "Reduce blended CAC by 15% while holding conversion rate flat" can be checked, by anyone, at the end of the 90 days. Write metrics specific enough that you and a skeptical stakeholder would agree on the outcome without a debate.
Building In a Review Checkpoint
A 90-day plan without a checkpoint at day 30 or 45 just runs blind until the end. Build in an explicit mid-point review where you compare actual results against the plan's assumptions and adjust — the plan is a starting hypothesis for the quarter, not a contract that must be followed regardless of what the data shows.
Conclusion
A growth plan earns its value from focus: one named constraint, the right modules applied to it, milestones you can check, and a checkpoint that lets you adjust before the 90 days are wasted on an assumption that turned out to be wrong.