Picture a twenty-person logistics software company in April. The active list reads: a pricing overhaul, a partner integration, two enterprise pilots, a CRM migration, a hiring push, and a rebrand that started in January. Every item has a champion. Every item has a slide. Nothing has shipped since February. The leadership team spends its Monday meeting re-ranking the list, which is the most comfortable thing they could possibly do, because ranking feels like deciding without any of the discomfort of deciding.
The more useful question is what to stop doing when a business has too many priorities competing for the same five days. Not what to rank higher. What to end.
This is a commitment problem wearing a productivity costume
Almost every article on this subject treats priority overload as a personal discipline issue: build a better list, block your calendar, learn to focus. That framing survives because it is flattering. It suggests the business is fine and the people are merely disorganised.
The evidence points somewhere less comfortable. Atlassian’s State of Teams 2024 found that 64% of knowledge workers say their team is constantly being pulled in too many directions, and 70% say it would be easier to make progress with fewer, more specific goals. That second number is the interesting one. People are not asking for more time or better tools. They are asking their leadership to commit to less.
The same research reports that executives estimate only 24% of their teams are doing mission-critical work. Read that from the other side: the people who set the priorities believe three quarters of the effort under them is not critical, and it continues anyway.
Work does not die of natural causes. In a business, nothing stops because it ran out of relevance. It stops because a specific person decides, on a specific day, that it stops. Absent that decision, projects simply persist, drawing salary, meeting time, and attention forever.
The test that exposes the real list
Write down everything currently considered active. Beside each item, answer two questions: when was it last meaningfully advanced, and who last consciously decided it should continue?
Anything that has not moved in six weeks and has no answer to the second question is not a priority. It is a zombie. It was approved once, and it has been coasting on that approval ever since. In a business that feels overloaded, zombies usually make up a third of the list, and killing them costs nothing except the mild embarrassment of admitting they were never going to happen.
Obligations are not priorities, and confusing them keeps your list long
Here is a distinction that shrinks most priority lists by half before any hard choices are made.
Payroll runs. Tax filings happen. Safety requirements are met. Contracts already signed with customers are honoured. Regulatory deadlines are hit. None of these are priorities. They are the floor. They do not compete with anything, they cannot be traded off, and ranking them against a rebrand is a category error.
Yet they routinely appear on the same list as ambitions, which produces a list nobody believes can be cut. Of course it cannot be cut. Half of it is legally mandatory.
Split the list into three piles and treat each one with a different instinct:
- Obligations. Never prioritised, never debated. Standardised, documented, automated where possible, and assigned to someone whose job is to make them boring. The goal is for obligations to consume as little leadership attention as the electricity bill.
- Bets. Deliberate, funded, time-boxed attempts to change the business. Few in number. Each has one owner.
- Everything else. Deleted. Not parked. Deleted.
The third pile is where the article gets uncomfortable, so let us go there.
What to actually stop
Stop maintaining a backlog you are never going to build
There is a genuine methodological argument here, and I will take a side. The conventional position is that you should keep a master backlog so no idea is lost and no stakeholder feels ignored, then re-rank it regularly. Basecamp’s Shape Up method takes the opposite view: work is chosen in six-week cycles, and pitches that are not selected do not go onto a central backlog at all. If someone still believes in an idea later, they re-advocate it in a future cycle.
I think the second approach is right for almost every business under a few hundred people, for a reason that has nothing to do with software methodology. A backlog is not a storage system. It is a guilt pile. Every review session, the same forty items get read, re-argued, re-scored, and re-deferred. The cost is not the storage. The cost is that finite leadership attention gets spent maintaining a museum of decisions nobody wants to make.
The fear is always the same: we will lose good ideas. In practice, good ideas are stubborn. If an idea matters, a customer will raise it again, a salesperson will complain again, or its champion will bring it back. Call it the resurrection rule. An idea that cannot survive one person remembering it and arguing for it a second time was never going to survive contact with a hard quarter.
Delete the backlog. Keep a short list of what you are actively doing and a cycle boundary at which anything can be proposed again.
Stop treating responsiveness as work
Atlassian’s 2024 research found that 65% of knowledge workers believe it is more important to quickly respond to messages than to make progress on top priorities. That is not a personal failing repeated by two thirds of the workforce. That is a culture accurately reporting what gets rewarded.
If your fastest responders get praised in team meetings and your slowest, deepest workers get asked why the project slipped, you have already told everyone what the real priority is. The fix is not a notification policy. It is changing what leadership publicly admires. Praise the shipped thing, in front of everyone, by name, and stop thanking people for replying at 10pm.
Stop meeting to find out what happened
Atlassian reports that in organisations with poor meeting cultures, people spend 50% more time in unnecessary meetings than they do making progress on high-priority work. Asana’s 2023 Anatomy of Work Global Index, which surveyed 9,615 knowledge workers, found leadership loses 3.6 hours a week to unnecessary meetings, and its accompanying release noted executives were 30% more likely than the average worker to miss deadlines because of too many video calls or meetings.
Notice who suffers most. The people setting priorities are the people whose calendars are most fragmented, which means the decisions that would reduce everyone’s workload are made by the group with the least uninterrupted time to make them.
One rule solves most of this. If a meeting contains no decision that could genuinely go either way, it is a document. Status is a document. Updates are a document. Reviews of things nobody intends to change are a document. Keep the meetings where a real fork exists and someone in the room has the authority to choose.
Stop running two of everything
Half of knowledge workers in Atlassian’s research say they have worked on a project only to find out later that another team was doing the same thing. Fifty-six percent say teams at their company plan and track work in different ways, which makes collaboration harder. Asana’s index puts leaders at 10 apps per day, says 62% of the workday is lost to repetitive, mundane tasks, and reports workers believe better processes would save them 4.9 hours a week. Atlassian estimates 25 billion work hours are lost annually to ineffective collaboration within the Fortune 500 alone.
Duplicate work is usually described as a coordination problem. It is better understood as a symptom of priority overload. When a company is running fifteen initiatives, no individual can hold the whole picture, so people stop checking and start starting. Two teams solving the same problem is what a long list looks like from the inside.
The stop here is concrete: one place where active work lives, one format for how work is planned, one owner per problem area. Not the best tool. The same tool.
Stop accepting work through channels nobody owns
Every unowned intake channel is a promise-making machine. A customer emails the founder directly and gets a yes. A salesperson agrees to a custom feature in a renewal call. A board member forwards an article and someone treats it as a mandate. None of these commitments passed through the process that decides what the company does, and all of them consume the same capacity.
Close the side doors. Requests arrive in one place, get a named owner, and get an answer that is allowed to be no. The founder who cannot stop saying yes in customer emails is not being responsive. They are quietly overruling their own strategy several times a week.
Stop starting anything without a kill date
A deadline says when something should be finished. A kill date says when you stop paying for it regardless of how far it got. Unscoped side projects survive because they have the first and not the second. Give every bet a fixed time budget and a date at which it either ships, gets explicitly renewed, or ends. Renewal must be an active decision made by a named person, not a default.
The right number of priorities is not a number
Plenty of advice insists on three priorities, or one per quarter. The appeal is obvious, and the discipline behind it is sound, but no serious evidence establishes a universal cap. A seasonal retailer, a regulated clinic, and a two-location restaurant group do not share a magic number. What the research supports is the direction, not the digit: fewer and more specific beats more and vaguer.
So count owners instead of priorities. A priority is real only when one named person can spend a substantial share of their week on it and has the authority to make decisions without convening a committee. If you have four people capable of owning something and three of them are half-consumed by obligations, your ceiling is roughly two live bets. Not three because three sounds balanced. Two because that is the capacity you actually have.
Then apply the make-room rule. Nothing enters the active list without a named subtraction. If a new idea arrives and nobody can say what it displaces, it has not been added to the strategy. It has been added to everyone’s evenings.
This is also where the honest version of tradeoff thinking lives. You are choosing under incomplete information, and waiting for certainty is itself a way of keeping everything alive. The practical skill is making a decision you can defend with the information you have and then committing to it long enough to learn something.
Where tools help, and where they quietly lie to you
The pressure right now runs toward adding. Gartner’s 2024 CEO survey, fielded among 416 CEOs and senior executives, found 62% named growth as their top business priority, up from 49% in the prior year. Thirty-four percent identified AI as the top theme of the next business transformation, against 9% for operations efficiency. Growth and AI are both expansion stories. Almost nobody’s top priority is subtraction, which is precisely why subtraction is where the advantage sits.
Work-management vendors have moved in the same direction. Atlassian now includes Rovo Search, Chat and Agents on every paid Jira plan, with Premium listed at $14.54 per user per month. Asana has pushed toward project intake, resource planning, goal management and AI products. The framing has shifted from personal productivity to organisational alignment, which is the correct diagnosis.
But note what better tooling does to an overloaded business. It makes tracking forty initiatives feel manageable. AI triage can summarise a backlog you should have deleted. Faster search can help you find a project nobody should be working on. The tool removes the friction that would otherwise have told you the truth.
Pricing structure is more revealing than feature lists here. Basecamp sells a Freelancer plan at $25 a month for up to three active projects, a Studio plan at $59 a month for up to ten, and an Unlimited Edition at $300 a month flat with 1 TB of storage. Charging by active project rather than by seat puts a small tax on sprawl. Per-seat pricing does the opposite: adding a fifteenth project costs nothing, so fifteen projects appear. If you want your tooling to support focus, choose a system where starting something has a visible cost.
One more thing tools do that rarely gets discussed. Vendors add priorities to your list without asking. Atlassian has said it will end sale of new Jira Data Center licences on 30 March 2026 and end Data Center support on 28 March 2029. For a company running a self-hosted setup, that is a migration project entering the roadmap on someone else’s schedule. Platform lifecycle dates belong in your capacity planning, not as a surprise in a future quarter.
My position on the underlying disagreement is simple. Tooling helps after commitment volume comes down. Before that, it industrialises the overload.
Replace the long list with a short cycle
Cutting once does nothing. The list regrows within a quarter, because the forces that created it are still running. What holds is a rhythm where the list is rebuilt from scratch at a fixed boundary and protected in between.
The Small Business Administration notes that a lean startup plan is typically one page and can take as little as an hour, and positions the format as a way to visualise tradeoffs and fundamental facts about the company. That is the wrong document to use once at launch and the right document to rewrite at every cycle boundary. One page forces omission. A thirty-page plan can hold every idea you have ever had, which is exactly why it never helps you choose.
Basecamp’s six-week cycle is a reasonable reference point rather than a law. What matters is that the boundary exists, that work is chosen for the cycle rather than accumulated indefinitely, and that the period inside the boundary is defended. The boundary is also the answer to the fear that focus becomes rigidity. You are not refusing to reconsider. You are refusing to reconsider on Tuesday because someone read something interesting. Anything urgent that is not an obligation waits for the boundary, which is usually weeks away, not years. If a business finds itself losing focus as it grows, the missing piece is nearly always this boundary rather than the goals themselves.
Within a cycle, the useful discipline is turning whatever you chose into something with dates, owners and visible checkpoints. That is a separate craft, and one worth doing properly once you have translated a goal into a short execution plan rather than a wish.
The part that will feel wrong
Cutting priorities is not painful because the items are bad. It is painful because the items are good. The integration would genuinely help. The rebrand would genuinely look better. The second product line might genuinely work. Killing obviously bad work is administration. Killing good work is strategy, and it feels like loss because it is one.
Every business that feels overloaded already knows which three things matter. What it lacks is permission to say out loud that the rest is not happening this year. Leadership’s job is to grant that permission explicitly, in writing, with names attached, so that nobody is quietly protecting a project they think is still alive.
A business does not become focused by choosing what matters. It becomes focused the first time it ends something that mattered.
