Strategic Priority

Past about twenty people, planning meetings change shape. Every team brings a proposal that makes sense from where they sit, and collecting those proposals onto one page produces more work than the company can carry. Marketing brings four, engineering four, sales four — twelve items, none of them objectionable, all approved.

Three months later the result is always similar. Twelve efforts are each about 70 percent done and nothing has shipped. Nobody was idle; one person was spread across three efforts. When someone asks for a status update, twelve lines come back, and reading all twelve still doesn't answer what the company accomplished this quarter. The cost isn't only the unfinished work — a person switching between three efforts loses time re-entering each one, so the same hours produce less than they would on a single track.

The cause was treating prioritization as putting things in order. Numbering items one through twelve feels organized while discarding nothing. Number twelve means "someday," and in practice someday is this quarter. Strategic priority is the correction: write down what was chosen and what will not be done, narrowing the stretch of the vision that gets worked on now.

Gather the candidates, cut them to around three, and actually move people and budget onto those three. Six steps.

1. Collect candidates. Team proposals, requests customers keep repeating, homework created by the competitive situation — all on one page. Nothing gets cut at this stage.
2. Rank them on shared criteria. Three are enough: does it move the company toward the vision, does it thicken a competitive advantage, and can it finish inside the period. Drop the third and the list fills with two-year projects that show progress every quarter and results in none of them. Applying the criteria in the room, out loud, matters as much as the criteria themselves — a proposal that nobody can argue for against a shared yardstick is usually one that only made sense to the team that brought it.
3. Cut to roughly three. The number comes from staffing. At a 45-person company, an effort needs five to eight people genuinely on it to move, which makes three the number that can run at once.
4. Write down what won't be done. This step is half the work. Publishing a "not this quarter" list alongside the chosen items lets the teams who were counting on those items adjust their own plans early, and it stops the quiet version of the same problem, where an item nobody mentioned again still has two people working on it.
5. Move the resources. Without resource allocation following, only the document changed. If the people who were working on something outside the list are still working on it, the list is a statement rather than a decision.
6. Attach one line per item saying how completion gets judged. "Improve search" can't be judged; "search returns results in under a second across 10,000 documents" can.

Pagedeck's priorities this quarter are three: enterprise permission controls, search speed, and first-week activation after signup. The published not-doing list holds a mobile app, a public API, and a new pricing tier. Nobody disputed that all three are needed — only that they aren't this quarter's work. Each of the three carries a named owner and a completion line, so the quarter can end with a yes or a no rather than a percentage.

The list does real work in three settings.

1. The quarterly planning meeting. It is where candidates get cut and where the reasons get recorded. Without the reasons, the same proposal repeats the same debate next quarter, and the teams that lost the argument have no way to tell whether the answer was "no" or "not yet."
2. On top of a goal-management framework. At a company running OKRs, this list is the raw material for the objectives. Write OKRs without priorities first and every team sets goals while nobody makes a company-level choice.
3. Building the budget. If the money allocated per period points somewhere other than the priorities, one of the two is untrue. The fastest check is whether budget still sits against items that aren't on the list — spending tends to keep its old shape long after a plan changes.

Its most frequent use is somewhere else, though: declining requests that arrive mid-quarter. When an important account asks for a feature and the person fielding it can answer from this list instead of deciding alone, the refusal becomes a company decision rather than a personal judgment.

"Three priorities this quarter — anything else goes into next quarter's candidate list."

Work that isn't on the list doesn't start now.

"Tell me in one line which of the three priorities this feature attaches to."

Asking a new proposal to fit inside the chosen set.

"The priorities changed but the budget didn't. So they didn't change."

Pointing out that the document and the resources disagree.

At Pagedeck, Grace sets the three before the quarter starts and publishes them company-wide. Mia brings the product candidates, while Ravi and Sofia bring the marketing and sales homework. The part that takes longest isn't deciding what goes in; it's explaining what comes out.

Last quarter the mobile app was the item cut. Requests for it had been steady, sales pushed back, and Grace settled it by making it the first candidate for the next quarter.

Reading priority as sequence is the most common error, since the word means order. But numbering and choosing are different acts. Number twelve items and all twelve stay on the list, and anything on the list gets started by somebody. Staffing splits twelve ways and the quarter ends with nothing finished. The point is not ordering the list but shortening it, which is why a well-made priority document has a longer not-doing section than doing section.

The other error is thinking a long list preserves flexibility. Keeping options open sounds safer when circumstances might change. But every open option is a judgment you have to make again each week. At a company with twelve candidates, each incoming request reopens the question of what is most urgent from scratch, and the time spent on that judgment comes out of the time available to do the work. A company holding three can answer the same request in five minutes, because the answer is already written down and the only question left is whether this request beats one of the three.