Strategies List Priorities. They Rarely Rank Them.
A strategy can be fully agreed and still decide nothing, because it names several good things without saying which one gives way when they collide.
A strategy that lists three good things has not chosen between them.
The document exists. It was worked on for a quarter, presented to the board, distributed to every function, and nobody in the building would say it is missing. It names the markets to enter, the margin to protect, and the speed the roadmap has to hold. Every sentence in it is something the leadership team actually believes.
And still, six months later, the same decisions come back. A regional pricing exception gets approved, reversed, and re-approved. A feature gets pulled from a release to protect a margin target, then pushed back in to protect a launch date, then pulled again. Nobody is being difficult. Each person making each call can point to the strategy and be right.
Three goods, no order
That is the tell. A strategy that names growth, margin, and speed as priorities has stated three things a company wants, not the order in which it will give them up. Any decision that only ever touches one of the three is easy, and the strategy handles it fine. The decisions that keep reopening are the ones that require trading two of them against each other, and on those, the document is silent.
Different rooms apply the same strategy and land in different places.
A regional sales lead reads "enter three new markets" and approves a discount to close a logo in a market the strategy names. A finance lead reads "protect gross margin" and blocks the same discount a month later, citing the same document. Both are following the strategy correctly. It told each of them what to protect and never told either of them which protection wins when protecting both is not possible.
This is why the pattern reads as inconsistency rather than disagreement. Nobody in either room thinks they are contradicting the strategy. Each is applying a faithful, partial reading of it, and the two readings only collide once they meet in the same account, the same release, or the same quarter.
The same document, opposite answers
A product organisation shows the identical shape. A roadmap commitment to ship on a stated date and a commitment to hold a security or compliance bar both sit in the same strategy, unranked. When a release is running late and the choice is between slipping the date or shipping with a known gap, the engineering lead and the commercial lead can each open the same slide and defend opposite calls. The strategy did not fail to anticipate the conflict. It anticipated it and declined to resolve it, which is a different thing from an accident.
A register closes a decision. It cannot close a trade-off the strategy left open.
The correction for a decision that keeps returning is usually an owner, a stated basis, and a trigger for reopening it, held in a register that is boring to check. That correction is right, and it is not enough here, because the register can faithfully record two decisions, made by two owners, each with a stated basis and a real trigger, that simply weighted the same unranked priorities in the opposite order.
A well-run register makes this visible sooner. It does not make it stop, because the register closes an individual choice; it has no answer for the question sitting underneath the choice, which is which of the strategy's stated goods actually outranks the others. Two owners can each close their decision correctly and still leave the organisation running two different strategies, both signed off, both traceable to the same page.
What a register can and cannot see
A register shows that a decision was reopened. It cannot show that the reason was structural rather than personal — that the same conflict will produce the same disagreement the next time it appears in a different account or a different release, wearing a name nobody has used before.
Ranking a strategy means naming what loses, and that is the part left out.
Naming three priorities costs nothing at the point of writing the strategy. Ranking them costs something immediately, in the room where the strategy is approved, because it means the board and the leadership team have to agree, in writing, which of their stated goods will bend first when two of them cannot both hold.
That sentence is uncomfortable in a way a list of priorities is not. "Margin bends before the launch date moves" is a real commitment, and it is one somebody will be unhappy about the first time it costs them a number. A list that says growth, margin, and speed all matter offends nobody in the room where it is approved, and that is exactly why it gets written that way.
The list is agreed unanimously because it commits to nothing. The ranking is argued over because it does.
Where the order actually lives
In practice a rank order does not need to cover every case, and rarely can. What it needs is enough of an order to answer the two or three trade-offs a company predictably faces:
- Which bends first when growth and margin cannot both hold this quarter.
- Which bends first when a launch date and a quality or compliance bar cannot both hold.
- Which function's number is allowed to miss so another function's does not.
A strategy that answers those three questions in one sentence each does more decision work than most strategies do in twenty pages of stated priorities, because it tells every room downstream what the last room already worked out for itself, badly and inconsistently.
Ranking priorities is strategy work, done before the room disagrees.
This is not a case for fewer priorities. A company legitimately wants growth and margin and speed, and a strategy that pretended otherwise would be lying about its own ambition. The gap is not in wanting three things. It is in never having said, out loud and in writing, which one gives way first.
That sentence is cheap to write before a live trade-off is on the table and expensive to write once it is, because by then two functions have a number attached to the outcome and neither wants to be the one who bends. Doing the ranking early is not a smaller version of the same argument; it is the same argument at a moment when it costs nobody anything to have.
That is the shape of the work we do at the strategy stage, alongside the trade-off work that turns a statement into a decision that can be carried: naming, before the quarter that tests it arrives, which of a company's stated goods actually comes first. The rest of decision architecture — ownership, registers, triggers — holds far better once that order exists to hold onto.
Continue