What documents should come out of a product strategy process?

What leaves the strategy room is a small set of documents, and each of them recommends rather than decides. The first-version definition, the roadmap with its criteria, the technical path, the handoffs to partners: every one ends 'subject to the team's judgment.' The facilitator recommends; the team decides. That sentence is not politeness. It is what keeps the accountability where the first whitepaper put it.

What is the difference between a recommendation and a decision in a startup?

What leaves the strategy room is a small set of documents, and each of them recommends rather than decides. The first-version definition, the roadmap with its criteria, the technical path, the handoffs to partners: every one ends 'subject to the team's judgment.' The facilitator recommends; the team decides. That sentence is not politeness. It is what keeps the accountability where the first whitepaper put it.

How should a product consultant or facilitator hand off strategy work?

What leaves the strategy room is a small set of documents, and each of them recommends rather than decides. The first-version definition, the roadmap with its criteria, the technical path, the handoffs to partners: every one ends 'subject to the team's judgment.' The facilitator recommends; the team decides. That sentence is not politeness. It is what keeps the accountability where the first whitepaper put it.

Documents that recommend · Product Strategy

§The question: what leaves the room, and in what voice?

§ is corrected and the open items have owners. Now someone writes it down, and how they write it decides whether the team owns the strategy or merely received it.

§ 12.1The set#

§Four documents, at most, sent within the week, each short enough to be read at a desk.

§The first-version definition. The bottom of the ladder, as chapter 9 described it: goal, scope in and out, use cases with acceptance criteria, appetite, technical path.

§The roadmap. The board with horizons, dependencies and promotion criteria, and the "can we do X to move Y" line for every item.

§The technical path. How the first version will be built, by whom, on what, with the trade-offs named. Usually written by whoever will build it, and reviewed by the owner.

§The handoffs. Whatever partners need: a brand partner, a development partner, an operations lead who will run the pilot. Each gets a page that says what is decided, what is open, and what they are being asked to do.

§The framework, the decision document and the ladder itself sit behind these as the record. The four above are what people act on.

§Every one of them ends with a version of the same sentence: subject to the team's judgment.

Weak voice	Strong voice
"The MVP is the customer profile and tier. Development begins on the 3rd."	"Recommendation: the first version is the customer profile and visible tier, for the reasons and with the evidence above. Subject to the team's judgment."
"Payments has been moved to month twelve."	"Payments is recommended for the twelve-month horizon, promoted when unaided check-in reaches five of six. The team may weigh the business case differently and should say so in the decision document."

§The difference is not tone. It is who is accountable. The first whitepaper placed product accountability with one named person on the team. A facilitator, a consultant, an advisor, or the co-founder who happened to write the strawman, does not hold that accountability, and a document that decides on their behalf quietly moves it. Marty Cagan's argument in Empowered is that teams own outcomes only when they own the decisions that produce them; a strategy that was handed down is a strategy the team will execute and not defend.

§ 12.3What recommending is not#

§It is not hedging. A recommendation is stated with full confidence and full evidence: this is what we should do, here is why, here is what would change it. Amazon's narrative memos recommend in exactly this register, and they are among the most decisive documents in business. Peter Drucker's definition of a decision as a commitment to action still holds; the recommendation is the proposal for that commitment, made as strongly as the evidence permits.

§Nor is it a way to avoid being wrong. The recommender is on the record. If the recommendation is followed and fails, the recommender was wrong and the document shows it. The team's ownership of the decision does not dilute the recommender's ownership of the advice.

§ 12.4Dates and owners#

§Each document carries a date by which the team will confirm, amend or reject it. A recommendation without a response date is a decision by default, which is the thing the sentence was meant to prevent. In the hospitality the four documents went out by email with a date the following week, and the team's confirmation, with two amendments, closed the strategy phase.

§ 12.5What comes out#

§Four documents that recommend, dated, with response dates, confirmed or amended by the team in writing. The ladder is now the team's. Part IV is how all of this fails, and the cases that show it working.

12. Documents that recommend