Pick five people off your team. Ask each one separately, with no warning and no preparation: what is this team for? One sentence.
Then line up the five answers.
I have run this exercise. The answers rarely match. Not five variations on one idea... five different ideas, pointing five different directions, from five people who sit within earshot of each other and ship the same code.
The numbers are worse than you expect
Donald Sull of MIT Sloan and Rebecca Homkes of London Business School surveyed 11,000 senior executives and managers across more than 400 companies. Only a third named their firm's top three priorities correctly. Given five attempts, around half managed to list the same single priority as their colleagues.
A later MIT Sloan Management Review analysis of 124 organisations found 28 percent of executives and middle managers responsible for executing strategy able to list three of their company's strategic priorities. One third of the leaders charged with implementing the strategy listed not a single one.
Read those again. These are senior people. Many of them sat in the room where the strategy got written. If they walk out of the building unable to repeat it, your engineers have no chance at all.
Your mission statement is doing nothing
Somewhere on a wall or a wiki page sits a sentence about empowering customers to achieve more through innovative solutions. Nobody has read it since onboarding. Nobody would notice its removal.
Mission statements fail for one dull reason. They get written to survive a review. Every stakeholder holds a veto. Anything specific enough to exclude someone's pet area gets softened until it fits everybody. What emerges offends nobody and directs nobody.
The waste is real. Gallup's August 2025 survey of 4,475 US working adults found 18 percent saying their current job has a purpose they believe in. The same research found employees who strongly agree their organisation's mission makes their job feel important report strong work purpose at 58 percent, against 16 percent for everyone else.
So mission matters enormously when it lands. Most of them never land.
A one-liner has to sting a little
Here is the difference between a mission statement and a one-liner. A one-liner excludes things.
Use this shape:
We [do this specific thing] for [these specific people] so [this changes for them]. We do not do [the obvious adjacent thing].
Examples I would accept from an engineering team:
- "We keep checkout working under load for shoppers on bad mobile connections. We do not own the catalogue."
- "We make new internal data queryable within a day of arrival, for analysts who do not write SQL. We do not build dashboards."
- "We carry the overnight pager for everything customer-facing so product teams sleep. We do not write features."
Each names a beneficiary, a change, and a boundary. Each one will irritate somebody in a planning meeting. Good. Irritation is evidence of content. A sentence nobody objects to has said nothing.
Three tests before you keep it
Test one: does it exclude? Read the second half aloud to the person most likely to lose out. Their face tells you whether you wrote a boundary or a pleasantry.
Test two: does it survive a Tuesday? Take a real ticket from last sprint, one you argued about. Apply the sentence. If it fails to settle the argument, it is decoration.
Test three: would a new starter act on it? Someone in week two, with no context and no relationships, reads your sentence and knows which of two requests to pick up first. Or they do not.
Most one-liners die on test one. Leaders write the first half with enthusiasm and then go vague on the exclusion, because the exclusion is where the political cost sits. Pay the cost. The whole value of the sentence lives in the second half.
The engineering bill for a missing sentence
The famous statistic here says 64 percent of software features are rarely or never used. I have repeated it myself over the years. Then I checked the source. It traces to Jim Johnson of the Standish Group, speaking at the XP 2002 conference, drawing on four internal applications at four companies. Mike Cohn walked through how thin the basis is. Four apps. 2002. Time to retire it, me included.
You need no borrowed number anyway. Open your own backlog and count three things. How many tickets originated with nobody outside your team. How many features shipped last year with usage you have never measured. How many migrations began under one rationale and continued under a completely different one.
Your number sits right there, and it belongs to you.
Teams without a one-liner keep working. They work on whatever arrives. The loudest stakeholder wins, every time, because every request looks equally valid with no sentence to measure it against. Prioritisation degrades into a popularity contest between people instead of a comparison between outcomes.
Saying no requires a stated reason. Without a one-liner, "no" sounds like laziness or empire protection. With one, "no" sounds like focus. Same word, entirely different politics.
Your roadmap is not a one-liner
Leaders push back on this with a link to the quarterly roadmap. Twelve initiatives, colour coded, owner names attached. Surely everyone knows the direction.
A roadmap lists what happens next. A one-liner explains why any of it earned a slot. Strip the roadmap of the sentence underneath and it becomes a list of things somebody once approved, which is precisely how teams end up defending work nobody wants any more.
OKRs land in the same trap. A team with four objectives and eleven key results has a scoreboard, not a purpose. Ask the same five people to recite the objectives from memory and watch the room study its shoes. The Sull research covers senior leaders with the same problem, and those people wrote the objectives.
One sentence, held in memory by everyone, beats a beautiful document nobody opens between planning sessions. Memory is the whole point. A purpose needing a lookup fails in the exact moment it matters, when someone gets asked for a favour in a corridor and decides alone.
Write yours this week
- Write your own version first, alone, in one sentence. No committee, no workshop, no facilitator with sticky notes.
- Ask each person on the team for theirs, separately. Separately matters... in a group, the first answer contaminates every answer after it.
- Put all the versions on a wall with the names stripped off.
- Argue about the boundary, not the adjectives. The "we do not" half is where the genuine disagreement hides. Wordsmithing the first half is avoidance dressed as work.
- Settle on one sentence. Then use it out loud in the next three planning meetings, applied to a real decision with a real loser.
Step five does all the work. A sentence nobody ever uses to kill a ticket is wall art.
Six weeks later, run the five-person survey again. Matching answers mean it took. Five different answers mean you wrote yourself another mission statement.
This one sits with you
Nobody delegates this. Not to the strategy team, not to HR, not to the offsite facilitator. Leadership shows up in what you refuse, and refusing needs a stated basis. Most of the leadership failures I have studied and written about over at Step It Up HR start in the same place... a leader who assumed everyone understood, and never once tested the assumption.
Go and ask the five people. Write down what you hear before you start defending anything. The gap between your sentence and theirs is the exact size of the problem you have been managing around without seeing.
What would your five say?