Anyone can take a percentage off a budget. The skill is doing it without quietly cutting the thing that was earning you money. Cost control is not a one-off purge before a board meeting; it is the ongoing practice of understanding what each pound buys, removing waste that no customer would miss, and pointing scarce resources at the work that actually limits your output.

The quick version

  • Cost control is the ongoing discipline of understanding and managing what you spend, not a one-time cut. Efficiency is getting more output from the same input. Resource planning is matching your people, money and capacity to demand, ahead of time.
  • You cannot manage costs you cannot see. Activity-based costing traces overhead to the activities that cause it, so you learn which products and customers actually make money, often a surprise.
  • Zero-based budgeting rebuilds the budget from nothing each cycle, justifying every line, instead of inheriting last year's number plus a few percent.
  • Efficiency has a catch: in any system, one bottleneck sets the pace. Saving time or money anywhere else changes nothing. Find the constraint first.

The idea in depth

The first mistake is treating cost as a single number to shrink. Two costs of the same size can be completely different decisions: one keeps a customer, the other keeps a habit. To tell them apart you have to see where money goes and what it produces, which is harder than it sounds, because traditional accounting spreads overhead across products using a blunt average (machine hours, headcount, revenue) that bears little relation to what each product actually consumes.

That blind spot is what activity-based costing (ABC) was built to fix. In their 1991 Harvard Business Review article "Profit Priorities from Activity-Based Costing," Robin Cooper and Robert Kaplan argued that companies should trace overhead to the activities that drive it, setups, inspections, order processing, and then to the products and customers that demand those activities. The finding was uncomfortable: a small share of products and customers generated most of the profit, while a long tail quietly destroyed it. The low-volume, high-touch order that looked fine under average costing was often sold at a loss once you counted the disruption it caused.

So the move is to stop asking "is this cost big?" and start asking "what activity causes it, and does the thing that activity serves earn its keep?" You do not need a full ABC system to use the idea, pick your three largest overhead pools, ask what actually drives each one, and trace them to the handful of products or customers responsible. Most leadership teams discover at least one product they should reprice, redesign or retire, and one customer whose business is costing them money to keep.

An honest limitation. Full activity-based costing has a poor track record of sticking. Many companies that adopted it later abandoned it because the data-gathering, surveying staff on how they spend their time, maintaining dozens of activity pools, was expensive, slow and irritating to maintain. Kaplan himself acknowledged this and, with Steven Anderson, proposed a lighter "Time-Driven Activity-Based Costing" (HBR, 2004) to cut the overhead. The lesson for a working leader: use ABC's logic, costs are caused, not averaged, without necessarily building the heavyweight machinery. A rough trace you actually maintain beats a precise model you abandon.

Rebuilding the budget instead of inheriting it

Knowing where money goes is half the job; deciding where it should go next year is the other half. Most budgeting is incremental: take last year's number, add or subtract a few percent, argue at the margin. The flaw is that it bakes in every past decision, good or bad, a programme funded for a reason that expired three years ago keeps its money simply because it had it last year.

Zero-based budgeting (ZBB) attacks that inertia. It was developed by Peter Pyhrr at Texas Instruments and set out in his 1970 Harvard Business Review article "Zero-Base Budgeting", influential enough that Jimmy Carter, then governor of Georgia, hired Pyhrr to apply it to the state budget. The premise is simple and demanding: every budget starts at zero, and every line must be justified from scratch each cycle, as if the activity were brand new. Nothing is entitled to its funding by precedent.

flowchart LR
  A(["Last year's budget"]) --> B{"Which approach?"}
  B -->|"Incremental"| C(["Add or subtract a few %
inertia kept intact"])
  B -->|"Zero-based"| D(["Start from zero
justify every line again"])
  D --> E(["Fund by current value,
not by precedent"])
Incremental budgeting defends the past; zero-based budgeting makes every line re-earn its place. Leaders Loop

So the move, even if you never run a full ZBB exercise, is to pick one or two cost categories a year and budget them from zero, travel, software subscriptions, agency spend, recurring projects, and make each owner argue for the money as though it were new. As McKinsey notes in "The return of zero-base budgeting," the real prize is not a one-time cut but a durable culture of cost ownership, where managers know their numbers cold and stop treating budget as an entitlement.

Where it goes wrong. ZBB is often mistaken for a cost-cutting weapon, and used that way it backfires: it is slow, demoralising to run across everything at once, and can starve long-horizon investments (research, brand, maintenance) whose payoff doesn't show up in this cycle's justification. Done well it reallocates money toward what matters; done as a blunt purge it just makes everyone defensive. Rotate it through the budget a slice at a time rather than imposing it on the whole organisation in one painful year.

Efficiency has one true target: the bottleneck

Here is where most efficiency drives waste their effort. The instinct is to make everything faster and cheaper everywhere. But in any system that turns inputs into output, performance is capped by a single constraint, the slowest, most-loaded step, and improving anything else changes nothing. Eliyahu Goldratt made this the spine of his 1984 business novel The Goal and his Theory of Constraints: an hour saved at the bottleneck is an hour gained for the whole system; an hour saved anywhere else is a mirage.

An hour saved at the bottleneck is an hour gained for the whole system. An hour saved anywhere else is an illusion.

This reframes resource planning. Instead of spreading budget and headcount evenly, or chasing 100% utilisation everywhere, which only piles up half-finished work in front of the constraint, you find the one resource that limits throughput and feed it. Goldratt's prescription is a loop: identify the constraint, get the most out of it before spending anything, subordinate everything else to keep it fed, then invest to lift it, and when it moves, find the new one. The constraint might be a machine, a specialist team, an approval step, or a single overworked person. Locate it, and you know exactly where your next pound of capacity should go.

flowchart TD
  A(["Identify the constraint
the one step that limits output"]) --> B(["Exploit it
get maximum from it, free"])
  B --> C(["Subordinate everything
keep the constraint fed, never idle"])
  C --> D(["Elevate it
invest to add capacity here"])
  D --> E{"Has the constraint
moved?"}
  E -->|"Yes"| A
  E -->|"No"| D
Goldratt's five focusing steps, spend on the constraint before anything else, then repeat when it moves. Leaders Loop

One caveat worth keeping honest about: the bottleneck lens is sharpest in flow-based work, manufacturing, fulfilment, ticket queues, anything with a measurable throughput. In knowledge work the constraint is real but harder to see and may be a person's attention or a decision rather than a station on a line. Use it as a question, "what one thing, if it had more capacity, would lift everything downstream?", rather than a formula.

A worked example

Take a mid-sized print-and-fulfilment business, call it Merrow, under pressure to cut 10% from costs. (Illustrative figures throughout; this is a teaching example, not real accounts.) The reflex is across-the-board: trim every team's budget by a tenth and call it discipline.

Instead, Merrow's leaders run the three lenses above. First, a rough activity-based trace on overhead reveals that a "premium" small-batch product line, flattering under average costing, actually consumes most of the setup, inspection and customer-support hours, and is sold at an illustrative loss of around £3 on every unit. That single insight saves more than any uniform cut: reprice it, and the loss becomes a margin without touching anyone else's budget.

Second, they zero-base two categories, software subscriptions and outside agency spend, and find roughly £80k a year of tools nobody could justify from scratch and a retainer that had outlived its original project. Third, they look for the bottleneck and find it: a single large-format printer that every premium job has to pass through, idle 15% of the day waiting on manual setup. An across-the-board cut would have trimmed the staff who keep that printer fed, making the constraint worse. Instead Merrow spends a little to streamline its setup, lifting throughput on the one machine that gates the whole operation.

The 10% target is met, but not by shrinking everything by a tenth. It is met by repricing the product that was losing money, cutting spend that wasn't earning its place, and investing at the constraint while protecting it from the cut. Same headline saving; a stronger business on the other side, instead of a weaker one.

Frequently asked questions

What's the difference between cost control and cost cutting?

Cost cutting is a one-off reduction, usually under pressure and often blunt. Cost control is the ongoing discipline of understanding what you spend, why, and what it produces, so you can keep value-creating spend and remove waste deliberately. Cutting treats all costs as equally disposable; control distinguishes the cost that keeps a customer from the cost that keeps a habit.

Do we need a full activity-based costing system to benefit from it?

No, and many firms that built full ABC systems later abandoned them as too costly to maintain. The valuable part is the logic: costs are caused by activities, not spread by averages. Trace your two or three largest overhead pools to the activities and customers that drive them. That rough exercise usually surfaces a loss-making product or customer without the machinery of a permanent ABC model.

Isn't zero-based budgeting just an excuse to slash budgets?

It can be misused that way, and it backfires when it is, starving long-horizon investments and demoralising managers. Done well, ZBB is about reallocation, not pure reduction: it forces every line to re-earn its funding so money moves toward what currently matters, not what mattered three years ago. The safe way to start is to rotate it through one or two cost categories a year rather than imposing it everywhere at once.

How do I find the bottleneck in my team's work?

Look for where work piles up and waits. In flow-based work it is the step with the longest queue in front of it; in knowledge work it is often a single approval, a specialist everyone depends on, or a recurring decision that stalls everything downstream. Ask: "if this one thing had more capacity, would everything after it speed up?" If yes, you have found the constraint, and where your next investment belongs.

Should I aim for 100% utilisation of my people and equipment?

No. Running every resource flat out only stockpiles half-finished work in front of the constraint and lengthens lead times. The Theory of Constraints view is that non-bottleneck resources should have slack; their job is to keep the constraint fed, not to look busy. Full utilisation everywhere is a vanity metric, not an efficiency one.

Related in the Toolkit

Cost control sits on top of how you read your numbers and how you plan ahead, you cannot manage spend you cannot see in the financial statements, and the budget is where these disciplines become a plan rather than a wish, which is why budgeting is the closest neighbour to everything here.

Where to go next