Playbook
How to write a three year plan for a consulting business
Asked for a three year plan and about to write an ambition instead: a dated future announcement, an FAQ, and a review structured so it can produce a no.
The methodThe working backwards brief, with a gate
This playbook is a method for writing a three year plan for a consulting business using Amazon's working backwards technique: a one-page press release dated three years out, written in the past tense with no internal language, an FAQ that answers the hard questions, and a review structured so that it can produce a no. It is for founders and chief executives of services businesses who have been asked by a board or an investor for a three year plan and are about to write an ambition instead. The method is Amazon's. The counterweights are mine, added because the subject is your own company rather than a feature, because a services plan has to describe a change in shape, in mix, in utilisation (utilization) and in who the customer is, rather than only a change in size, and because a document about your own future has a way of being quoted back as a promise.
A three year plan for a services business usually fails in one of two ways. Either it is a spreadsheet, revenue growing at a rate chosen because it looked reasonable, with no account of what the company would have to become for the numbers to happen. Or it is a vision statement, all adjectives and no dates, which nobody can be held to and nobody can argue with. Neither tells the board what decision it is being asked to make. A consulting business has a particular version of this problem: its capacity is people, its recurring revenue is usually a small share, and it cannot grow much by working the same people harder, so the plan has to say what the company will stop being as well as what it will become.
Amazon's six steps come first, credited and unchanged: the customer, the problem, the solution, the experience, the proof, and the FAQ. Run on a company, the press release announces something that has happened to the business three years from now: the customer it serves, what it is known for, and the one number a stranger would need. The FAQ answers what an investor, a customer and an employee would ask, including the questions you would rather not. Then the three counterweights. First, every forward number is labelled as such, on the page, so it can never be cited later as history. Second, the reversibility test: before writing a brief at all, ask whether the decision it points to can be undone, because working backwards is a tool for irreversible choices and a wasted week for reversible ones. Third, the gate: the review is structured so that it can produce a no, with the criteria for a no written down before the brief is read, and it ends in a recorded decision rather than an agenda item.
What is deliberately not in it: a valuation method, a revenue forecasting model, or any claim that a brief written this way came true. The output is fiction, useful fiction, and the playbook says so on every page that carries a number. There is also no template for an exit itself: no buyer list, no process, no advice on advisers. The brief is about what the company becomes. Whether anyone buys it is a separate question and a separate playbook.
I learned it by being asked. After outside investment came into the business that became DevOpsGroup, the board wanted a three year plan, and my first draft was an ambition with a spreadsheet attached. The version that worked was written as a press release about a company that did not exist yet, with a date on it and a customer named by type, and the FAQ beneath it was harder to write than the release. Reading it aloud to the board produced the argument the spreadsheet never had, because the release made a claim about what the company would be known for, and people disagreed with the claim. The company that page described did not come to exist in that form. Some of it happened, some of it did not, and because the numbers had been labelled, nobody could later pretend the page had been a forecast. The discipline held, and the fiction stayed fiction.
The playbook is in development, with the press release and FAQ template alongside it. Joining the waitlist means you hear when it is ready, and it lets me ask what your board has actually asked you for. Nothing is for sale yet.
What you will get
Who it is for
Questions
- What should a three year plan for a consulting business contain?
- A dated statement of what the company has become, written as a one-page press release in the past tense; an FAQ that answers what an investor, a customer and an employee would ask; and a review that can say no. Every forward number is labelled on the page. A spreadsheet on its own is not a plan; it is an extrapolation.
- Is Amazon's working backwards method suitable for a services business?
- Yes, with counterweights. Amazon wrote it for products and features. Run on a company it needs a reversibility test, so you only use it for decisions that cannot be undone; a label on every forward number; and a gate so the review can produce a no. The six steps are Amazon's and are credited as such.
- How do you stop a three year plan being quoted back as a promise?
- Label it. Every forward number carries the word forecast or the word fiction beside it, on the page, and the review's recorded decision states what was decided rather than what was predicted. The playbook treats this as a rule, not a courtesy.
- What is the point of a consulting business plan if the output is fiction?
- The argument. A dated press release makes a claim about what the company will be known for, and a board can disagree with a claim in a way it cannot disagree with a growth rate. The value is the decision the argument produces, recorded, with a no available.