Build a 90-day plan with weekly moves

This prompt asks a model to act as a hands-on startup advisor and produce a growth plan for the next quarter. You give it the context of your business, and it returns a schedule broken into three 30-day phases. Each phase names the single number that matters most, three specific weekly actions, the revenue you should expect, and a short list of activities to cut. It is built for founders and small teams who want a tight, fast-moving plan rather than a long strategy document.

When to use it

Use it at the start of a quarter, or whenever you feel busy but not growing. It works best when you already know your main constraint, such as weak lead flow, low conversion or too much manual work. If you are still figuring out what is broken, spend a day on that first. The prompt is also useful before a planning session, since it gives you a draft to argue with rather than a blank page.

How to use it

1. Fill in the four inputs honestly: what the business does, current monthly revenue, the one bottleneck that hurts most, and team size. Vague inputs produce vague plans.
2. Decide how aggressive you want the targets. Add a line stating your risk appetite and any hard limits, such as cash on hand or hours you can work.
3. Paste the prompt with your inputs into your model of choice.
4. Read the first phase closely. Check that each weekly action is something a person on your team could start on Monday, not a theme like improve marketing.
5. Push back on anything that needs a new hire, a new tool or a budget you do not have. Ask for a version using only current resources.
6. Ask the model to list the assumptions behind its revenue numbers, then correct the ones that are wrong for your market.

What a good result looks like

A strong answer is specific enough to schedule. Each phase should name a metric you can measure weekly, such as qualified calls booked or average order value, and tie it to a number. The actions should read like tasks: rewrite the pricing page, call the ten largest lapsed customers, cut the weekly report nobody reads. Revenue impact should be a range with the reasoning behind it, not a single hopeful figure.

The cut list matters as much as the plan. A good answer removes work, not just adds it. If every phase is packed with new projects and nothing is dropped, ask again and say you have no spare hours.

Expect the model to get some details wrong about your market. Treat the output as a first draft you edit. The value is the structure: one metric, three weekly moves, a number and a cut, repeated three times. Keep that shape and replace the content with what you know to be true.