Growth at All Costs Was an Operating Model

The sentence arrives fully formed and usually in the same week across a whole sector. This is the year of profitable growth. It appears in a board pack, then in an all-hands, then in the language managers use with their teams, and within a month everyone can say it. Almost nothing about how the business runs has changed, and the plan for the quarter is still the plan that was written against the old target.

What makes the switch feel simple is that it is stated as a change of preference. The company used to want volume and now it wants margin, and preferences can be changed by announcing them. But growth at all costs was never a preference. It was a specification, and the organisation was built to it: the headcount, the channel mix, the pricing floor, the length of the sales cycle the team is staffed to absorb, the customers who were acceptable to sign.

The old target is still in the building

A company does not hold a growth strategy the way it holds an opinion; it holds it in the shape of its costs. Those commitments were made over years, by people acting correctly under the instruction they had, and they are load-bearing. Changing the instruction leaves every one of them in place and now unexplained.

The gap between the announcement and the machine is where the next two quarters get lost. Numbers move slightly, because discretionary spend is cut and hiring pauses, and both are read as evidence that the shift is working. Then the effect stops, because everything easy has been done and what remains is structural.

Read a cost base as an archive rather than a budget and it becomes legible. Each line was added to solve a problem that the previous target created, and it stays after the target is gone because nobody has been asked to explain it against the new one.

• A support organisation sized for a customer population that was acquired on price and asks more of it.
• A field sales team staffed for a deal size that the incentive plan encouraged people to chase downwards.
• An implementation function that exists because the product was sold ahead of what it could do unassisted.
• A brand and demand spend whose true job was to keep the top of the funnel wide enough for a conversion rate nobody wanted to defend.

Cuts land on the newest commitments first

When the instruction changes, the reduction that follows tends to fall on whatever was added most recently, because recent additions have the least political protection and the clearest owner. That is unrelated to which commitments the old target actually created. The cost of growth at all costs is concentrated in the oldest structural decisions, and those are the ones a cost programme is least likely to reach.

So the company removes the layer it added last year, keeps the machinery it built three years ago for a customer it no longer wants, and reports a leaner operation with the same underlying economics.

Mix is decided upstream of the quarter in which anybody notices it. A customer signed under a volume instruction was signed at a price that cleared, into a segment that answered, with commitments made in the room to get the deal closed. Those terms persist for the life of the account, and the account is now part of the base the new target is measured against.

A mix problem always presents as a margin problem, one reporting period too late.

This is where the shift is most often mistaken for a pricing exercise. Raising list price does not touch the installed base, and the installed base is the part carrying the cost. The economics of a company that grew at all costs are set by who it agreed to serve, not by what it charged them, and the second is far easier to change than the first.

Retention reads differently under a new target

The same accounts also distort the retention picture that the new plan depends on. Retention measured across a base assembled under a volume instruction blends customers who would have bought anyway with customers who were bought, and reports one number for both. The blend was fine when the job was to grow the total. It is misleading the moment the job becomes to grow the profitable part of it.

Untangling that takes a quarter of unglamorous work and produces no announcement. It is also the only version of the shift that survives contact with the following year.

Margin is the last thing to move, which makes it a poor instrument for steering. A company that manages the transition by watching margin will be a year into the change before the reporting tells it anything it did not already know, and by then the decisions that produced the result are too old to revisit.

The earlier signal is the payback period on newly acquired revenue: how long a customer signed this quarter takes to return what it cost to sign. It is a claim about the future rather than a record of the past, which is what makes it useful and what makes it uncomfortable. It also moves within one cycle of a real change in who the company sells to.

What a payback number exposes

A payback figure that will not hold still is not a measurement problem; it is the shape of the old target showing through. When the number swings by segment, by channel, or by which rep closed the deal, the company is still acquiring on the instruction it says it has replaced, and the average is hiding it.

There is a decision underneath this that deserves to be made deliberately rather than absorbed, and it has the structure of the reversible and irreversible split. Cutting a campaign is reversible within a quarter. Leaving a segment, retiring a price point, or dismantling the function that served a customer type is not, and those are the moves that actually change the economics.

The companies that make the change stick treat it as an operating question rather than a financial one. They name the customers they will no longer pursue, in language specific enough that a salesperson can act on it on a Tuesday. They rebuild the incentive plan before the quarter opens rather than after it disappoints. They decide which commitments made under the old target are being kept on purpose, and say so.

Narrower, not lighter

None of that is a cost programme, and all of it is subtraction — the same discipline that governs what a growing company can still read when it has added faster than it has removed. Profitable growth is not a lighter version of growth. It is a narrower one, and the narrowing has to be done by hand, account by account and line by line.

Where the reading gets done

This is the work we do in consulting engagements: reading the cost base as a record of past instructions, testing the mix against the target the company now says it has, and naming the irreversible moves before they are made by drift. It sits inside the wider argument we keep making about growth discipline — that a company's real strategy is whatever its costs and incentives are still built for, regardless of what the current slide says.