Executive summary: A turnaround is not an improvement program run faster. It is a different sequence entirely: stabilize cash, diagnose honestly, fix the structural cause, then grow. Companies fail at turnarounds because they attempt those four in the wrong order, usually starting with growth initiatives while the cash position deteriorates underneath them. This guide covers the phases, the first 90 days, how to assess an inherited team, what to cut and what to protect, and how to tell whether it is working before the financials confirm it.
- What is a business turnaround?
- Do you need a turnaround or an improvement plan?
- What are the four phases of a turnaround?
- Why does cash come before everything else?
- How do you diagnose what is actually wrong?
- What do you do in the first 90 days?
- How do you assess an inherited leadership team?
- What do you cut and what do you protect?
- How do you sequence quick wins against structural fixes?
- How do you communicate a turnaround internally?
- How do you know it is working before the financials say so?
- How long does a turnaround take?
- What are the most common turnaround mistakes?
- Business turnarounds: operator FAQ
- About the Stagnation Assassin
What is a business turnaround?
A turnaround is a structured intervention to reverse sustained financial decline, run in a specific order: stabilize cash, diagnose the true cause, fix it structurally, then rebuild growth. It differs from continuous improvement because the timeline is compressed and the failure mode is running out of money before the fix lands.
The distinction matters more than it sounds. Continuous improvement assumes you have time. It assumes the business survives whatever pace you set, so you can sequence work by convenience, build consensus, pilot carefully, and expand. A turnaround assumes none of that. The clock is the binding constraint, and every week you spend building agreement is a week of cash you will not get back.
I have led turnarounds inside Berkshire Hathaway, Illinois Tool Works, Whirlpool, and JBT Marel, and the pattern that separates the successful ones from the failures is almost never the quality of the strategy. It is the sequence. Companies that stabilize cash first and diagnose second survive long enough for their strategy to matter. Companies that lead with a growth initiative because it is more inspiring tend to run out of runway while the initiative is still in pilot.
Here is the part nobody says out loud. Most struggling companies do not have a strategy problem. They have a focus problem wearing a strategy costume. They are doing forty things reasonably well, three of which determine whether the business lives, and nobody has been willing to name which three. The turnaround is largely the act of naming them and killing the rest.
Do you need a turnaround or an improvement plan?
You need a turnaround when the trajectory, not just the current result, is negative and when normal management action has already failed to reverse it. If performance is disappointing but stable and cash is not deteriorating, you need an improvement plan. If cash is shrinking and the trend line points at a wall, the answer is different.
The honest test is a trajectory question rather than a performance question. Plenty of businesses perform below potential for years without ever being in danger. That is an improvement problem, and treating it as a turnaround imposes trauma the situation does not require. Conversely, a business that still looks profitable on paper can be in genuine peril if working capital is quietly consuming its cash faster than earnings replace it.
Signals that you are past improvement and into turnaround territory:
- Cash is deteriorating regardless of reported profit. Profitable companies fail. Cash-generating companies rarely do. If those two numbers are diverging, believe the cash.
- Normal management action has already been tried and did not work. If the obvious fixes have been attempted by competent people and the trajectory did not change, the problem is structural rather than executional.
- The organization has stopped being surprised by bad news. This is the cultural signal, and it is the most reliable one. When missing a target produces no reaction, the organization has normalized decline.
- Every function is individually defensible and the whole is failing. Each department hits its metrics, the company still loses money, and nobody can explain the gap.
- Improvement effort is spread evenly across the business. Even distribution of effort is a reliable indicator that nobody knows what actually drives results.
That last one deserves emphasis because it is diagnostic. When I walk into an operation and find improvement resources allocated roughly proportionally across every function, I know before looking at a single financial statement that leadership cannot name the constraint. Even allocation is what an organization does when it does not know where the leverage is, and it is enormously expensive.
What are the four phases of a turnaround?
Stabilize, diagnose, restructure, and grow. Stabilization buys time by protecting cash. Diagnosis identifies the structural cause rather than the symptoms. Restructuring fixes that cause, which usually means removing things rather than adding them. Growth comes last and only after the underlying economics work, because growing a broken model accelerates the failure.
Phase 1: stabilize
The objective is time, not improvement. Stop cash outflow, secure liquidity, halt discretionary spending, and freeze commitments that are not yet contractual. Nothing in this phase fixes the business. It simply guarantees the business is still there when the fix arrives.
Phase 2: diagnose
Find the structural cause. Not the symptom everyone complains about, and not the explanation the previous leadership team has already settled on. This phase requires measuring reality rather than reading reports, and it is where most turnarounds are quietly decided.
Phase 3: restructure
Fix the cause. In practice this is usually subtraction: killing products, exiting customers, consolidating facilities, removing layers, and stopping initiatives that consume capacity without producing return. Addition comes later and costs money you do not yet have.
Phase 4: grow
Only once unit economics work. Growing a business whose underlying model loses money on each transaction makes the loss larger, faster, and more difficult to reverse. This is the phase everyone wants to start with, and starting with it is the most common way turnarounds fail.
The order is not stylistic. Each phase creates the conditions the next one requires. Diagnosis without stabilization gets interrupted by a liquidity crisis halfway through. Restructuring without diagnosis cuts the wrong things, which is worse than cutting nothing because it destroys capability you needed. Growth without restructuring scales a broken model.
Why does cash come before everything else?
Because cash is the only resource whose exhaustion ends the company outright. Profitability problems are survivable for years. Liquidity problems are survivable for weeks. Stabilization does not improve the business, it buys the time required to improve the business, and without that time nothing else you do matters.
The first move is always visibility. Most struggling companies have a monthly reporting rhythm and no reliable weekly view of cash. Build a thirteen week rolling cash forecast in the first fortnight, updated weekly, showing receipts, disbursements, and the resulting balance. This single artifact changes the conversation more than any analysis, because it converts a vague sense of pressure into a specific date.
The second move is stopping outflow that has not yet become contractual. Discretionary spend, uncommitted capital projects, new hiring, travel, consulting engagements that have not begun. None of this is strategic and all of it is reversible, which is exactly why it goes first. You are not trying to be clever in this phase. You are trying to extend the runway.
The third move is usually the largest and the least obvious: working capital. In most struggling manufacturers, a substantial amount of cash is sitting in inventory that protects nothing. When I helped scale a custom manufacturing business through a period where working capital became the binding constraint, restructuring inventory policy released roughly 4.2 million dollars while production volume rose 25 percent. That cash was already inside the business. It was simply allocated to parts that could be bought next day rather than to the ones that actually stopped production.
A thirteen week rolling cash forecast, updated weekly, changes the conversation more than any strategic analysis. It converts a vague sense of pressure into a specific date, and a specific date is what makes an organization willing to do things it has refused to do for years.
A caution on this phase. Stabilization is genuinely necessary and it is also where inexperienced turnaround leaders do permanent damage, because cutting is visible, fast, and rewarded. Cut discretionary spending hard. Be extremely careful about cutting capability, and do not touch the constraint. More on that distinction below.
How do you diagnose what is actually wrong?
By measuring rather than asking. The organization will offer you a consensus explanation for its own decline, and that explanation is usually the symptom everyone can see rather than the cause nobody has measured. Go to where the work happens, measure what actually limits results, and be prepared for the answer to contradict the received wisdom.
There is a specific reason the internal explanation is unreliable, and it is not dishonesty. Organizations converge on explanations that are politically survivable. An explanation that blames the market, the economy, a competitor, or a decision made by someone who has already left is comfortable for everyone still present. An explanation that identifies a specific structural failure implicates whoever owns it. Consensus explanations are selected for comfort, not accuracy.
What actually works is embarrassingly simple and rarely done. Walk the operation. Measure with a stopwatch rather than pulling standards from a system. Count inventory physically rather than trusting the database. Ask operators what stops them, because they usually know and have usually stopped being asked. I have identified the true limiting factor in a business in three days using nothing more sophisticated than observation and arithmetic, in situations where the leadership team had been confidently wrong about it for years.
Three diagnostic questions carry most of the weight.
What actually limits output or revenue right now? Not what is inconvenient. What is genuinely capping the result. In operations this is usually a specific process. In commercial businesses it is often a conversion step rather than a lead volume problem. When I worked with a services business convinced it needed to double marketing spend, the real constraint was quote conversion, and fixing the quote process delivered more revenue than the marketing increase would have, at roughly an eighth of the cost.
Where is effort going that produces nothing? Look at where people spend time and where improvement resources are allocated. In most struggling businesses, a large majority of improvement effort is pointed at things that cannot change the result. That misallocation is not incompetence. It is the predictable outcome of never having identified what matters.
What is the business unwilling to say out loud? Every struggling organization has one. A product line everyone knows loses money. A customer that consumes more than it pays. A leader nobody will confront. The diagnosis is often already known and simply unspeakable, and part of the outsider’s value is being able to say it.
What do you do in the first 90 days?
Weeks one to two build cash visibility and stop discretionary outflow. Weeks three to six diagnose by measuring the operation directly. Weeks seven to ten deliver two or three visible wins to establish credibility. Weeks eleven to thirteen commit publicly to the structural changes. The purpose of the first 90 days is earning the authority to do the hard part.
Weeks 1 to 2: visibility and the bleeding
Build the thirteen week cash forecast. Freeze discretionary spend, uncommitted capital, and non-critical hiring. Meet the people who actually do the work, not just their managers. Say very little about strategy, because you do not yet know anything worth saying and premature pronouncements are expensive.
Weeks 3 to 6: measure reality
Go to the operation. Measure the things everyone believes and check whether the beliefs survive contact with data. Expect substantial gaps between reported and actual figures on cycle times, availability, inventory, lead times, and customer profitability. Those gaps are where the opportunity lives.
Weeks 7 to 10: two or three visible wins
Deliver something the organization can see and did not think was possible. This is not window dressing. In a turnaround, credibility is a working capital of its own, and you spend it later on decisions that will be unpopular. Pick wins that are fast, visible, and genuinely connected to the diagnosis rather than easy wins that teach the wrong lesson.
Weeks 11 to 13: commit publicly
State the diagnosis, the plan, and the specific changes, including the unwelcome ones. Ambiguity at this stage is read as weakness and generates far more anxiety than bad news does. People can work through a hard truth. They cannot work through not knowing.
One note on pace. The instinct in the first week is to act decisively to demonstrate control. Resist it for the specific category of structural decisions, because a confident structural decision made on week-one information is frequently wrong and always expensive to reverse. Move fast on cash. Move fast on discretionary spending. Take four to six weeks on diagnosis, because that is the decision everything else depends on.
How do you assess an inherited leadership team?
Assess on two axes: capability and candour. Capability is whether they can do the job the business now requires, which may differ from the job they were hired for. Candour is whether they tell you inconvenient things without being cornered. The second predicts turnaround performance more reliably than the first.
Most people evaluate inherited teams on competence alone, and competence is the easier variable to read. But a highly capable executive who manages information is more dangerous in a turnaround than a moderately capable one who tells you the truth immediately, because in a compressed timeline you are making decisions on what you are told. Bad information delivered confidently costs you weeks you do not have.
Practical tests that work.
Ask a question you already know the answer to. Not as a trap, but as a calibration. How someone handles a question where the honest answer is unflattering tells you more in thirty seconds than a quarter of observation.
Watch what happens when they are wrong in public. Do they update, or defend? A turnaround requires constant revision as diagnosis improves. Someone who cannot be publicly wrong will slow every decision cycle they participate in.
Ask what they would do with full authority. Capable people who have been blocked have a detailed answer ready and often a surprisingly good one. People who have accepted decline do not have an answer at all, and that absence is itself informative.
Look at who they have developed. Leaders who build capability leave evidence in the people around them. This is the hardest signal to fake and the most predictive of whether they can execute the rebuild phase.
On timing, make these decisions faster than feels fair and slower than feels urgent. I have made both errors. Moving too fast, I removed a leader in week three whose difficult reputation turned out to be the result of being the only person willing to raise problems. Moving too slowly, I kept someone for two quarters because the team liked them, while their function quietly failed to execute anything. Six to eight weeks of direct observation is usually enough, and it is worth the discipline of writing your assessment down at week two and testing it at week eight rather than trusting a continuously revised impression.
What do you cut and what do you protect?
Cut activity that consumes capacity without producing result. Protect the constraint, the customer relationships that generate genuine margin, and the small number of people who actually hold institutional capability. The distinction is between cutting cost and cutting capability, and the two look identical on a spreadsheet.
This is where turnarounds do permanent damage. Across-the-board cuts are administratively simple and strategically catastrophic, because they remove the same percentage from the function that determines your output and the function that produces nothing. The result is a smaller version of the same broken business, with less capability to fix itself.
What consistently deserves cutting:
- Products and customers that consume more capacity than they return. In most portfolios a meaningful tail of products generates negligible margin while consuming disproportionate operational attention. Killing that tail releases capacity immediately and costs almost nothing in real revenue.
- Improvement initiatives aimed at things that cannot change the result. I have walked into businesses running dozens of active improvement projects where almost none touched the actual limiting factor. Halting them is free capacity.
- Layers that exist to coordinate other layers. Coordination cost grows faster than the organization does, and in decline it is rarely revisited.
- Reporting that nobody acts on. One plant I visited tracked 47 metrics on its production board. The plant manager could not name three that would tell him whether yesterday was a good day. That is measurement obesity, and it consumes real hours.
What deserves protection even under severe pressure:
- Anything that touches the constraint. Cutting maintenance, staffing, or material availability at the process that limits your output reduces the company’s total earning capacity to save a small cost. It is the most expensive saving available.
- The people who hold undocumented capability. Every operation has a handful of people who know how things actually work. Losing them in a cost reduction is an unrecoverable error, and they are frequently not senior.
- Customer relationships with genuine margin. Distinguish these from large customers, which is not the same thing. Revenue concentration and margin concentration are often in different places.
The refrigeration division I took over was losing 175 million dollars a year, and the consensus internally was that we needed new equipment, more people, and more floor space. That was wrong in every particular. We had enormous idle capacity sitting in the wrong places while the actual limiting factor went unmanaged. We cut the annual losses by more than half without adding a single machine or a dollar of capital, and eventually reached profitability. Almost all of it came from stopping things rather than starting them.
How do you sequence quick wins against structural fixes?
Run both simultaneously, but understand that they serve different purposes. Quick wins buy credibility and organizational belief. Structural fixes buy actual recovery. A turnaround built only on quick wins stalls in month six when the easy gains are exhausted and nothing fundamental has changed.
The sequencing rule I use is that every quick win should be a visible instance of the structural argument. If your diagnosis is that the business improves everything except the thing that matters, then your first quick win should be a dramatic improvement at the thing that matters, delivered fast. Now the win is not just a win, it is evidence for the case you are about to make about how the whole company should operate.
Quick wins that teach the wrong lesson are worse than no quick wins. Cutting travel budgets produces a real number and teaches the organization that the turnaround is about austerity. Recovering a large block of capacity from an existing asset teaches that the problem was focus rather than resources, which is the lesson you actually need to install.
On structural work, start it in parallel from week seven even though it will not produce results for months. Structural fixes have long lead times, and a turnaround that waits until quick wins are exhausted before beginning structural work creates a visible plateau at exactly the moment when belief is most fragile.
A turnaround built only on quick wins stalls in month six. The easy gains exhaust, nothing structural has changed, and the organization concludes the effort failed. Start structural work in parallel from week seven, even though it produces nothing visible for months, because its lead time is the real constraint on recovery.
How do you communicate a turnaround internally?
Say the hard thing early, specifically, and once. Ambiguity generates more fear than bad news, because people fill silence with worse scenarios than reality. State the situation, the diagnosis, what will change, and what will not, then repeat it consistently rather than softening it over time.
The instinct to protect people from bad news is understandable and counterproductive. Everyone in a declining business already knows something is wrong. Withholding the specifics does not spare them anxiety, it converts a definite concern into an indefinite one, and indefinite concern is what drives your best people to start taking recruiter calls. The people you can least afford to lose have the most options and will move first.
Four practices that work.
Be specific about what is not changing. This is more stabilizing than any reassurance. If the core business, the primary site, or a particular team is not at risk, say so plainly and early. Uncertainty spread evenly across an organization paralyses all of it.
Explain the reasoning, not just the decision. People will execute a decision they disagree with if they understand the logic. They will quietly resist one they cannot make sense of, and in a compressed timeline quiet resistance is fatal.
Do not promise a timeline you cannot control. Turnarounds slip. Committing publicly to a recovery date you miss costs you more credibility than admitting uncertainty would have.
Change the metrics people are measured on, visibly. Communication that is not backed by changed incentives is decoration. If you tell an organization that focus matters and then keep paying bonuses on the metrics that produced the sprawl, the bonus structure wins. It always wins.
That last point is the one I underestimated most badly, more than once. I assumed that showing people compelling data would change behaviour. It did not. People understood the argument intellectually and rejected it emotionally, because two decades of training told them otherwise, and because their compensation still rewarded the old behaviour. Changing what gets measured and rewarded is not an administrative detail that follows the strategy. In a turnaround it substantially is the strategy.
How do you know it is working before the financials say so?
Financial statements lag by a quarter or more, so track leading indicators instead: cash trajectory against forecast, output from the limiting resource, decision velocity, and whether bad news is surfacing faster. The last one is cultural and it is the earliest reliable signal that the turnaround is taking hold.
Four indicators worth watching weekly.
Cash against forecast. Not the absolute number, the variance. A forecast that is becoming more accurate means the organization is developing real visibility into its own operations, which precedes every other improvement.
Output at the limiting resource. If you have correctly identified what constrains the business, its output is the closest thing to a real-time proxy for company performance. It moves weeks before the financials do.
Decision velocity. Measure the time from a decision being raised to being made. Struggling organizations have slow decision cycles, often because approval structures have accreted over years. One appliance manufacturer I worked with required seventeen executive signatures for a product launch. Restructuring that into three tiers by reversibility cut launch timelines by 40 percent and improved decision quality, because executives were finally spending attention on decisions that actually required their judgment.
How fast bad news travels. In a declining organization, problems surface late and pre-packaged with explanations. When people start bringing you problems early and unresolved, the culture has shifted, and that shift precedes financial recovery reliably enough that I treat it as the primary indicator.
How long does a turnaround take?
Stabilization takes weeks. Diagnosis takes four to six weeks. Structural change takes two to four quarters to show in results. Full recovery to sustainable performance typically runs 18 to 36 months. The pace is set less by operational complexity than by how quickly leadership will change what people are measured and paid on.
Some reference points from work I have led. A refrigeration division losing 175 million dollars annually was cut to roughly half that loss and then to modest profit, without capital investment. A retail equipment manufacturer went from 48 million to 60 million in revenue while profit improved from 2 million to 10 million. An industrial scales and packaging business tripled profit from 6 million to 20 million over three years. A custom manufacturer scaled from 50 million to 67 million in revenue in 26 months.
What varies most across those is not the operational difficulty. It is organizational willingness. The fastest recoveries happened where leadership changed performance measures early and accepted the political cost. The slowest happened where leadership wanted proof of concept before changing the metrics, which creates a deadlock: the metrics prevent the behaviour that would generate the proof.
Three factors compress the timeline reliably: a genuine cash crisis that removes the option of delay, a leadership team willing to be visibly wrong, and a diagnosis specific enough that people can act on it without further interpretation. Three factors extend it: distributed decision rights with no clear owner, a culture where raising problems is punished, and any form of parallel improvement program that competes for the same resources.
What are the most common turnaround mistakes?
Five recur: leading with growth before the economics work, cutting across the board instead of by contribution, diagnosing from reports rather than observation, keeping the old performance metrics, and treating the turnaround as a project with an end date rather than a permanent change in how the business runs.
Mistake 1: leading with growth
The most common and the most expensive. Growth initiatives are inspiring, they generate goodwill, and they feel like leadership. Applied to a business whose unit economics are broken, they increase the rate of loss and consume the cash that stabilization was protecting. Fix the economics, then grow.
Mistake 2: across-the-board cuts
Administratively simple, strategically indefensible. Cutting every function by the same percentage removes capacity from the constraint at the same rate as from functions with slack. You end up with a smaller version of the same broken business and less ability to repair it.
Mistake 3: diagnosing from the conference room
Reports describe what a system was designed to capture, which in a struggling business is frequently the wrong thing measured optimistically. I have seen reported availability of 94 percent turn out to be 67 percent under observation, and standard cycle times off by more than half. Go and look. It takes days and it changes everything downstream.
Mistake 4: keeping the old metrics
Organizations announce new priorities while continuing to measure and reward the old ones. This produces visible compliance and invisible resistance, and the incentives win every time. If you are not prepared to change what people are paid on, do not bother announcing a new direction.
Mistake 5: treating it as a project
Teams execute the turnaround, results improve, the team disbands, and eighteen months later performance has drifted back because the underlying management habits never changed. A turnaround should end with a permanently different operating rhythm, not with a completion announcement.
The mistake I have made most often is underestimating political resistance to changes I could prove were correct. Data does not persuade people to accept something that threatens their standing, their identity, or their bonus. I now spend considerably more time on the change effort and considerably less on refining the analysis, because the analysis is usually right long before the organization is ready to act on it.
Business turnarounds: operator FAQ
What is the first thing to do in a business turnaround?
Build cash visibility and stop discretionary outflow. A thirteen week rolling cash forecast, updated weekly, converts vague pressure into a specific date and is usually more decision-changing than any strategic analysis. Stabilization does not improve the business, it buys the time required to improve it.
How long does a business turnaround take?
Stabilization takes weeks, diagnosis four to six weeks, and structural changes two to four quarters to show in results, with full recovery typically running 18 to 36 months. The pace depends less on operational complexity than on how quickly leadership will change what people are measured and rewarded on.
What is the difference between a turnaround and continuous improvement?
Continuous improvement assumes you have time to build consensus, pilot carefully, and expand gradually. A turnaround assumes the opposite: the clock is the binding constraint, and the failure mode is running out of cash before the fix lands. Same tools in many cases, completely different sequencing and tolerance for delay.
Should you cut costs first in a turnaround?
Cut discretionary spending immediately and be extremely careful about cutting capability. Across-the-board reductions remove capacity from the function that limits your output at the same rate as from functions with slack, producing a smaller version of the same broken business. Cut by contribution, and never cut the constraint.
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 VP of Global Product Strategy. 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 the speaking page or connect with him on LinkedIn.
Next step: a turnaround diagnostic
Your business does not have forty problems. It has two or three that determine whether it recovers, and nobody has been willing to name them. Book a 20 minute turnaround diagnostic and I will help you separate the symptoms everyone talks about from the structural cause nobody has measured. Start the diagnostic here.