Advisory Note
What a Margin Crisis Is Actually For
A margin crisis hands you permission to change almost anything. Most owners spend it on the smallest thing on the page.
- A cash crisis and a structure crisis look identical on the P&L. They need opposite treatment.
- Cut headcount in a structure crisis and margin recovers. For two quarters.
- The crisis is the asset. It is the only window where you can change anything.
- Cut, redesign, redefine. Cutting first closes the other two doors.
Wrong Move!
Diagnosis: Why the obvious move is the wrong one
It starts in a meeting, and the meeting always looks the same.
An owner. The finance manager, or whoever passes for one. A P&L on the table.
Margin has been sliding for four or five quarters. Revenue may even be holding, which is what makes it confusing. Cash is tighter than the P&L suggests, because it always is.
And around the twenty-minute mark, somebody says the sentence that ends every one of these meetings.
We need to look at manpower.
It's not an unreasonable instinct. Look at what else is on that page. Rent is contracted. Materials are priced by someone upstream. The financing is the financing. Of every line on an SME P&L, manpower is close to the only one that moves because you decided it should.
Under pressure, you reach for the thing that moves.
The instinct isn't the problem. The problem is that it answers a much smaller question than the one your business is actually asking.
Same Symptoms, Different Animal
A cash crisis is a liquidity problem in a business that works. A large customer defaults. A project overruns. A season doesn't arrive. The shape of the business is sound; it has just run out of runway to prove it. Cutting cost is the right treatment. It buys time, conditions normalise, and the business works again — because it always did.
A structure crisis is a different animal. You're running the business in a shape it grew into years ago. Since then the revenue mix moved. Customers changed what they buy and how they buy it. A channel that was a third of revenue is now a tenth, or the reverse. New work got absorbed by adding people, because adding people was faster and there was margin to absorb it. Roles accumulated. Handoffs multiplied. Nobody redrew anything, because nothing broke badly enough to force it.
In a structure crisis your cost base isn't too big. It's pointed at the wrong things.
And here is the trap. On the page, the two look identical.
- What happened
- A one-off shock — a customer default, a project overrun, a season that didn't arrive
- What the P&L shows
- Margin down, cash tight
- What cutting cost does
- Buys the time to recover
- What it needs
- Liquidity and time
- What happened
- The business changed shape; the organisation didn't
- What the P&L shows
- Margin down, cash tight
- What cutting cost does
- Buys two quarters, then costs you the recovery
- What it needs
- Redesign, then resize
The Cut That Costs You
Four things happen. Reliably.
You cut by cost, not by contribution
Under pressure, cuts follow the salary line, because that's where the savings are. The people who go are the expensive, experienced ones. Often the only ones who know why anything is done the way it's done.
The work doesn't leave
It redistributes onto whoever is left, usually the people already carrying the most. Nothing visibly fails in the first month. That's the trap.
The damage arrives late
Quality, responsiveness and error rates degrade over a quarter or two, in ways the P&L quietly blames on the market.
You lose the ability to fix anything
The people who would have led the redesign have gone, or are covering three jobs and firefighting.
And margin does improve. That's what makes this so hard to argue with in the room. It improves for two quarters, sometimes three. Then it relapses.
Only now you're smaller, more fragile, and holding fewer options than when you started.
The second round of cuts is always harder than the first. There is almost always a second round.
Your Crisis Is an Asset
The turn: The golden opportunity that expires
Here is the half that gets missed. It's the more important half.
In a normal year, how much can you actually change? Long-serving staff resist it. Family shareholders ask why. The bank prefers continuity. Customers have expectations. Everyone has a reason to leave things alone, and you rarely have a reason strong enough to override all of them at once.
A margin crisis dissolves every one of those objections at a stroke. For two or three quarters, you can propose almost anything and be listened to. Roles can be redrawn. Products can be dropped. Customers can be let go. Pricing can change. How the business makes money goes back on the table for the first time in a decade.
That window is the most valuable thing your business will hold this year. And it is perishable.
Most owners spend it on a 15% headcount reduction.
Three Doors. You Can Only Open One First.
What to do: Cut, redesign, or redefine
Cut
Reduce the cost of the business as it is. What you get: a smaller version of the same business.
Redesign
Rebuild how the work runs around what the business actually sells today. What you get: the same business, run properly. Margin recovers and holds.
Redefine
Rethink what you sell, to whom, how you deliver it and how you charge. What you get: different economics. This is where growth comes from.
Almost every SME cost programme stops at the first rung. A good number of consultants stop at the second.
The third is where the crisis pays for itself. And the third is only available while the mandate lasts.
Notice the sequencing. You can't redesign around people who have already left. You can't redefine a business you've just hollowed out. Cutting first doesn't only damage the business. It closes the two doors that were worth more.
To be clear: redesigning and redefining don't avoid cuts. They usually still end with a smaller headcount. The difference is that you know which roles and which people, the shape holds afterwards, and you haven't spent the mandate before working out what to spend it on.
Five Questions Most Owners Never Ask
Not abstraction. Not a two-year transformation. Five questions, answered properly with real numbers, over eight to twelve weeks.
Why eight to twelve? Long enough to get real numbers out of the system — cost by customer and by job usually has to be built, not looked up — and to test the answers against the people who actually run the work. Short enough that the mandate is still alive when you finish.
Fire the wrong customers
Cost-to-serve almost never matches revenue. It's common to find that a meaningful share of customers consume more than they contribute — the demanding ones, the slow payers, the ones whose orders are small and bespoke. You rarely see it, because the P&L reports in total, not by customer. Letting the wrong customers go is one of the fastest growth moves available to an SME. And in a normal year it's unthinkable.
Stop selling one-offs
Project work versus recurring work. One-off supply versus a service contract. A product versus the outcome it produces. This changes the quality of revenue, not just the quantity — and a business with recurring revenue meets the next downturn on completely different terms.
You probably own too much
What you own, what you partner, what you outsource. Most SMEs carry a vertical integration inherited from a stage they left long ago — capabilities built for a volume or a mix that no longer exists.
Your price is a fossil
Cost-plus pricing was set when the cost base made sense. If the cost base no longer makes sense, neither does the price. Pricing model is often the real constraint on margin, and it's almost never examined during a cost exercise.
You may be in the wrong business
More often than owners expect, the genuinely profitable business turns out to be a small activity running quietly alongside the large one that consumes all the attention. The growth question isn't "how do we grow". It's "which of these two businesses are we in".
These five aren't the whole list, and they aren't meant to be. Whatever answers you arrive at, two tests sit underneath all of them.
Does it scale?
Not whether you can sell more. Whether the next dollar of revenue costs less to serve than the last one did. A business that grows without that is buying itself more work.
Does it lay a foundation?
Most SMEs here are not playing a three-year game. They are family businesses meant to outlast whoever is running them now. A cost exercise answers this quarter. These questions answer the decade.
AI Won't Save a Broken Model
AI belongs in this conversation. Just not where it usually gets put.
Applied to the business as it runs today, AI makes the existing model cheaper. Worth having. Not a strategy. A faster version of a model that has stopped working is still a model that has stopped working.
Applied after you've answered the five questions, it becomes something else entirely. It changes what is economically possible. Customers who were too small to serve profitably become serveable. Work that only justified a senior person's time becomes routine. Service levels that needed headcount stop needing it.
Order matters. Decide what business you're in, then ask what technology makes that business possible. Do it the other way round and you get expensive tooling bolted onto a model nobody examined.
Six Questions You May Not Want to Answer
They aren't subtle, and that's the point. The diagnosis is usually sitting right there, available to any owner willing to answer honestly.
Tap the ones that are true of your business.
- Has your revenue mix changed more in the last three years than your organisation chart has?
- Can you say what each of your major cost lines actually buys you — per customer, per product, per job — rather than in total?
- Which of your customers would you not take on again today, and what share of revenue do they represent?
- How much of your team's week goes to work that exists because of how you're organised — rework, chasing, re-entering, waiting for approvals — rather than because a customer asked for it?
- If your three most senior people were unavailable for a month, what would stop?
- The last time margins recovered, did they recover because something changed, or because a good quarter arrived?
Nothing here is recorded or sent anywhere. It's for you.
Two different problems hide in those answers, and it's worth being precise about which is which. Structure is how the work is organised — roles, processes, who does what. Model is what you sell, to whom, and how you charge for it. Redesign fixes the first. Redefine fixes the second.
If the honest answers point at structure, cutting headcount buys you two quarters and costs you the recovery. If they point at the model, it costs you rather more than that.
None of this says headcount should never come down. Sometimes it must, and delaying that decision has killed as many businesses as taking it badly. The argument is about what you do first — and what you're willing to put on the table while you still can.
Some of these questions come up again in the survey at the end of this piece. If you've answered them honestly here, you've already done most of the work.
Nobody Has Asked You
The ask: What we don't know, and are collecting
One consequence of everything above is unavoidable. Redesign the work or redefine the business, and what people do changes. Roles shift. Work moves. Some jobs stop existing in their current form and others come into being. That isn't a side effect of the exercise — it is the exercise.
Which turns out to be exactly what Singapore's workforce development funding exists to support: redesigning jobs so that people are redeployed rather than retrenched, with SMEs supported at up to 70% of qualifying cost.
Most owners never ask, because they assume the support is for training and for technology. It doesn't occur to them that reorganising the business itself is the fundable part.
We Don't Have the Numbers. Yet.
Everything above is a position, formed over two decades of sitting in those meetings. It is not a finding.
Here's what nobody has: any Singapore-specific data on what SME owners actually did when margins came under pressure, and what happened next. Not just whether they cut. Whether anyone used the moment to change the business, and whether that worked better.
Plenty of adviser folklore. No numbers.
So we're collecting them.
The Singapore SME Margin Survey
What happened to your margins. What you changed — your costs, your organisation, or your business model. What happened afterwards.
Run in partnership with 
Ledgen, whose client base across Singapore's SME sector is what makes a study of this kind possible.
The survey closes on 23 October 2026. Findings are published on 19 November and presented that morning at a closed briefing in Singapore. Everyone who responds receives the report and an invitation to the room.
Over the past three years, what has happened to your gross margin?
Improved significantly Improved slightly Held roughly steady Declined slightly Declined significantlyTap one. The remaining nine questions take under four minutes.
Individual responses are never published or attributed — only aggregate findings appear in the report. Your email is used to send you the findings and the briefing invitation, and nothing else.
One commitment. If the data contradicts everything above — if the businesses that cut first turn out to have recovered better — we'll publish that, and say so plainly. There's no point running the survey otherwise.
And if the survey isn't the right conversation — if you read those six questions and recognised your own business in them — talk to Lex directly.
Somewhere this week, a meeting like the one at the top of this page is happening. The question isn't whether manpower comes up. It's what gets decided in the twenty minutes after it does.