The first time I had to make this call myself, the list in front of me had eleven items on it and every one was real. A machine that needed replacing, a customer bleeding margin, a planning system nobody trusted, a supplier on terms we could not afford, and seven more. Nothing on the list was wrong. The list was simply longer than the year.
What I did not have at the time was a rule for choosing, and I picked the largest number, which is what most people do. It was a defensible choice and it was the wrong one, for a reason that took me most of that year to see clearly. The rule below is what I use now. The full sequence it produces in a manufacturing setting is in how to turn around a manufacturing plant. This is the rule underneath it, which is what you need when your situation does not match the canonical order.
What should you fix first?
Fix first whatever is removing your options fastest, not whatever is worst. Sequence by reversibility rather than by size. A large problem that will still be exactly as fixable in six months can wait. A small one that forecloses three other choices while you look at it cannot.
This sounds like a subtle reordering and it is not. Applied honestly it produces a different first move than the size rule does in most turnarounds, because the things that foreclose options are rarely the things that dominate the P&L. Cash runway forecloses. A covenant test forecloses. A supplier’s exclusivity window forecloses. A key person deciding whether to stay forecloses. None of those is likely to be the biggest number on your list.
The reason the rule works is that a turnaround is a sequence of choices made with incomplete information, and the value of any position is partly the choices it still contains. Spending month one on your largest problem while a smaller one quietly closes two doors is a net loss even when the large problem improves.
For a turnaround that fixed the product before the marketing, see my podcast episode Chip Bergh’s Levi’s turnaround sequencing.
Why doesn’t the biggest problem go first?
Because size and urgency are different properties, and only one of them is time-sensitive. A structural cost problem worth several margin points will be worth the same several points in March. It is not going anywhere, and neither is your ability to attack it.
The objection here is obvious and worth answering directly: surely the biggest problem is where the return is. It is. The rule does not say the biggest problem is unimportant, it says it goes second. Reversibility determines order. Size determines effort once the order is set.
There is a second reason, which is about credibility rather than arithmetic. Large structural problems take two to four quarters to show a result, and in most turnarounds you do not yet have the standing to spend two quarters on something invisible. The reversibility rule tends to produce early moves that are fast and verifiable, which buys the room you need for the slow work.
The third reason is the one I got wrong that first year. Working the largest item first tends to consume your only scarce resource, which is the attention of the two or three people capable of executing anything. Point them at a nine-month structural problem in month one and every fast, reversible improvement available to you goes unmade, because the people who would have made them are in a working group.
Where does the standard order break?
The canonical order breaks whenever your business is not short of cash, which is more often than the literature assumes. Cash, then visibility, then decisions, then margin, then cadence is correct in a crisis. In a business with two years of runway, starting with cash release is solving a problem you do not have.
Run the reversibility rule instead and the order comes out differently. With no cash pressure, the fastest-closing option in most businesses is people: the two or three individuals who hold the operating knowledge and are currently deciding whether the new regime is worth staying for. That is a genuinely irreversible loss on a short clock, and it outranks a margin problem worth ten times as much.
The second common break is the acquired business, where the closing option is usually an integration decision with a contractual date attached. And the third is a business facing a customer transition, where the reversible-looking choice to defer a conversation becomes irreversible the moment the customer’s own planning cycle closes.
The pattern across all three is that the closing item is usually a decision somebody else is making, not a condition inside your business. That is what makes them easy to miss. Operators scan their own operation for problems and find plenty, and the things that actually remove options tend to sit at the boundary: a customer’s planning cycle, a supplier’s fiscal year, a lender’s review date, a person’s own timeline. Scan the boundary first, then scan inside.
What if two things are both irreversible?
Take the one with the shorter clock, and if the clocks are the same, take the one whose failure is harder to recover from. Two irreversible items with identical timing is rare enough that when it happens you should check whether you have actually described one problem twice.
Where it is genuine, the tiebreak is asymmetry of consequence. Losing a key person and losing a supplier contract might both close in six weeks, but one of them can be replaced with money and the other cannot. Rank by what money cannot buy back.
And say out loud that you are making a tradeoff. The failure mode when two urgent things collide is to attempt both at forty percent, which reliably produces two partial outcomes rather than one complete one. Naming which one you are letting slip, and telling the person who owns it, is worth more than the extra effort you would have split across them.
How do you know a stage is done?
A stage is done when it stops removing options, not when it is finished. That distinction is what keeps the sequence moving, because most stages are never actually finished and waiting for completion is how a turnaround stalls in its first phase.
Cash stabilization is done when cash is no longer the binding constraint on your choices, which is well before the working capital work is complete. Measurement is done when the numbers are trustworthy enough to act on, not when the reporting is good. The decision layer is done when escalations come back inside a day, even though the authority structure will need attention for another year.
The practical test is to ask whether the next stage is now possible. If it is, move, and leave the current stage running in the background with an owner attached. Sequential turnarounds fail far more often from over-completing stage one than from moving too early.
What does the order look like in practice?
Six steps, applied to your own list rather than to a template. The output will usually resemble the canonical order in a crisis and will diverge from it noticeably when there is no crisis, which is the point of having a rule rather than a list.
- Write the list. Everything, unranked, one line each. Ten to fifteen items is normal.
- Mark what has a clock. A date, a renewal, a decision somebody else is making, a person deciding. Most items will have none.
- Rank the clocked items by how soon they close. Not by size. This is your first tier and it is usually two or three things.
- Rank the rest by speed of return. What moves in 30 days, then 90, then a year. This is your second tier.
- Check the second tier for hidden clocks. Structural work often has a dependency you have not noticed, usually a capital cycle or a contract.
- Put growth last, deliberately. It is the one item that reliably fails when placed anywhere else.
That last step deserves its own sentence. Growth initiatives sit on top of an operating model and inherit it, so a growth push launched before the model changes produces a larger version of the business you were trying to fix. The Harvard Business Review work on growth outside the core argues that the adjacent move works when the core capability genuinely transfers. My reading of the same evidence from an operating seat is narrower: it transfers when the core is healthy, and a business in a turnaround does not yet meet that condition.
A limitation worth naming. This rule assumes you can see the clocks, and in the first fortnight of a new role you cannot. Several of the items that will turn out to be time-critical are invisible to you on day one, which is why the two-week diagnostic exists and why the ranking should be redone at day 30 rather than treated as settled. A sequence built on incomplete visibility and then defended for a year is worse than no rule at all.
About the Stagnation Assassin
Todd Hagopian is a Fortune 500 transformation executive who has generated $3B in shareholder value across Berkshire Hathaway, Illinois Tool Works, Whirlpool, and JBT Marel, where he serves as Global VP of Product Strategy and Marketing. Known as The Stagnation Assassin, he is the author of two published books: The Unfair Advantage: Weaponizing the Hypomanic Toolbox and Stagnation Assassin: The Anti-Consultant Manifesto. His blog is published in 15+ languages and read by operators worldwide. Bring him to your stage via his speaking page or connect with him on LinkedIn.
If nothing on your list has a clock, that is a finding rather than good news. A business with no closing options is usually not in a turnaround at all, it is stagnating. Send me your list and I will run Stagnation Genome scoring against it.
See also: When to Judge a Turnaround, and When Not To, I Just Inherited a Business. Now What?