How to Decide Whether to Discontinue an Unprofitable Product

The product manager has an announcement, and judging by the triumphant expression, someone has already ordered a trophy for killing our worst product.

Our least profitable product is apparently being discontinued before anyone has established whether it is actually losing us money.

Naturally, there’s a slide.

There’s a red margin percentage, an arrow pointing down, and the kind of confidence usually reserved for people who haven’t opened the underlying spreadsheet.

The initiative is called ‘margin optimization,’ which sounds considerably more sophisticated than ‘we may have just fired a product that was paying the rent.’

Before anyone starts ordering commemorative cupcakes, I’d like to ask one question.

What costs actually disappear when the product does?

Because if the answer is ‘almost none,’ we’re about to make the financial statements look tidier while the business earns less money.

A red number is a perfectly good reason to investigate, not a permission slip to stop thinking.

Here’s how I’d decide whether to discontinue a product without confusing accounting allocations with cash savings.

First, Tell Me What ‘Unprofitable’ Means

A product can lose money on a fully allocated P&L and still generate positive cash contribution.

Those statements aren’t contradictory.

They’re measuring different things.

Gross margin subtracts product costs included in cost of goods sold.

Contribution margin focuses on revenue less costs that vary with the relevant decision.

A product-level operating profit report may then subtract allocated rent, executive salaries, shared software, corporate finance, or other overhead.

I want the fully allocated report, too; I just don’t want it masquerading as a decision model.

The trouble starts when somebody prints the bottom line in red, circles it, and calls the meeting finished.

I want to separate three categories: costs that vanish immediately, costs that disappear only after a deliberate restructuring, and costs that remain no matter what we do.

The distinction isn’t academic.

It’s the difference between removing a loss and removing a customer contribution while keeping the same expensive building.

A Hypothetical Product That Looks Like a Disaster

Imagine a manufacturer with a product line generating $2 million in annual sales.

All numbers in this example are hypothetical, not results from a real company.

Its annual variable manufacturing, freight, and sales costs total $1.3 million.

The product therefore contributes $700,000 toward other expenses and profit.

Finance allocates $900,000 of plant and corporate overhead to it.

The product report shows a $200,000 operating loss.

Naturally, the red ink has everyone reaching for the emergency exit, even though nobody has checked whether the fire is real.

But suppose only $150,000 of that allocated overhead would actually disappear if we stopped selling the product.

The remaining $750,000 is rent, salaried supervision, systems, and shared support that the company would still pay.

If we discontinue the line with no replacement sales, we lose $700,000 of contribution and save $150,000.

Annual profit declines by $550,000.

Wonderful.

We’ve cured the reported product loss by making the company $550,000 poorer every year; somebody alert the awards committee.

The product’s $200,000 allocated loss was real as a reporting result.

It simply wasn’t the incremental economics of the decision.

The Calculation I’d Put in Front of Management

I would build a keep-versus-discontinue bridge, with every assumption traceable to an owner.

Start with revenue that would disappear.

Subtract truly avoidable variable costs.

Add fixed costs that can actually be eliminated, and specify when those savings arrive.

Then include shutdown costs, inventory write-downs, contract penalties, and any effect on sales of related products.

In our example, the keep case delivers $700,000 of annual contribution.

The discontinue case saves $150,000 of avoidable fixed cost.

That makes discontinuation $550,000 worse per year before one-time exit costs or cross-selling effects.

I would not include the $750,000 of overhead that simply moves to other products after the allocation formula changes.

Moving a cost to another spreadsheet tab does not make it disappear, although apparently it does make the presentation look better.

I realize that sounds obvious.

It becomes considerably less obvious when the product review is scheduled for Friday at 4:45 and everyone wants the red cells to go away.

What If We Can Use the Capacity for Something Better?

Now the decision gets interesting.

Suppose dropping the product releases machine hours and warehouse space that could support a new product.

The replacement is expected to generate $900,000 of annual incremental contribution after its variable costs.

It also requires $100,000 of additional annual dedicated support.

The existing product would forfeit $700,000 of contribution.

Discontinuing it saves $150,000 of genuinely avoidable fixed costs.

Replacing it adds $800,000 of net annual contribution after the new support cost.

The net annual benefit of switching is therefore $250,000: $800,000 plus $150,000 minus $700,000.

That’s a different recommendation.

And it comes with different questions.

Is the replacement demand contracted, or merely the most optimistic line in a sales pipeline?

Will the equipment actually be available, or does changeover consume half the theoretical capacity?

Are the replacement margins supported by current quotes and labor standards?

If the new product fails to materialize, do we still have the old customers?

A plan that requires six optimistic assumptions to become attractive is not a plan I’d approve without a downside case.

It’s a wish list wearing a contribution-margin badge and hoping Finance won’t ask for references.

The Four Options I’d Actually Compare

Option one: keep the product as-is.

That can be rational if it generates contribution, uses otherwise idle capacity, and has no better alternative.

It may not be glamorous, but glamour has yet to pay a supplier invoice.

Option two: repair the economics.

Raise prices, remove costly customizations, reduce scrap, consolidate shipments, or renegotiate materials.

A modest improvement in contribution might outperform an expensive shutdown.

Option three: harvest the product.

Stop pursuing low-value new orders, maintain profitable customers, reduce complexity, and let demand decline in a controlled way.

This is especially relevant when contractual obligations or customer relationships make an immediate exit costly.

Option four: discontinue and redeploy.

I like this only when the avoidable savings and credible alternative use of resources beat the contribution we’d surrender.

None of these options should win because somebody changed the bar chart from red to green and declared victory.

They win because the cash economics work.

The Customer Problem Nobody Wants to Put on the Slide

Products rarely exist in perfect isolation.

A low-margin component might be the reason a customer buys a much more profitable service.

A specialty SKU might secure a broader distribution agreement.

An unprofitable entry-level offer might support upgrades, renewals, or a larger basket.

I’m not arguing that every loss-making product deserves a sentimental rescue.

I’m arguing that we should measure the relationship before we break it.

I’d ask the commercial team to identify customers who buy the product alongside other offerings.

Then I’d estimate what share of the related contribution is genuinely at risk.

If discontinuation causes us to lose another $200,000 of contribution elsewhere, our original $550,000 annual disadvantage becomes $750,000.

That extra exposure deserves a place in the decision model.

It does not deserve to be hidden in the footnotes under ‘strategic considerations.’

When somebody tells me customers will simply switch to another SKU, I’d like evidence.

Customers have an irritating habit of making their own decisions, generally without consulting the assumptions tab.

Don’t Ignore the First-Year Cash Bill

Even a sensible long-run exit can create a brutal short-run cash problem.

Suppose discontinuation requires $120,000 in severance and contract termination payments, plus a $180,000 inventory write-down.

The write-down is an accounting charge; its cash impact depends on when inventory was purchased and whether it can be sold or salvaged.

The termination payments, however, require cash.

I’d build a monthly transition forecast showing final customer shipments, collections, supplier settlements, shutdown payments, and any replacement-product ramp.

Then I’d compare the first-year and steady-state economics separately.

In the replacement scenario, the modeled $250,000 annual benefit does not mean $250,000 appears in the bank on January 1.

If the replacement takes six months to ramp, the first-year result may be far smaller or negative.

Timing matters.

Especially when the presentation has decided to put ‘annualized savings’ in 28-point font and the cash costs in something resembling microscopic legal disclosure.

The Scorecard I’d Use Before Signing Off

I want a decision table, not a hundred-tab spreadsheet that requires a guided tour, three passwords, and a small prayer.

For each option, show incremental revenue, variable costs, avoidable fixed costs, replacement contribution, exit cash costs, working-capital changes, and the downside scenario.

Put a named owner beside every important assumption.

Have Operations confirm capacity and labor changes.

Have Sales validate customer retention and replacement demand.

Have Accounting explain how the reported margin will change without pretending that allocation changes are savings.

Have Treasury verify the cash timing.

Then give management a recommendation with explicit conditions.

If the replacement product fails to hit its committed volume, what happens?

If we retain the line, which pricing or process changes must be delivered by which date?

And if we exit, who owns customer communication and inventory disposition?

A decision isn’t finished because the CFO signs the slide.

It’s finished when someone owns what happens next.

When I’d Kill the Product

I would discontinue a product when the contribution we give up is outweighed by costs we can truly avoid, credible contribution we can earn elsewhere, and other measurable benefits, after accounting for transition costs and risk.

I would not discontinue it merely because an allocation report shows a loss.

The allocation report is a clue, not a court verdict.

And it is certainly not a termination letter.

Sometimes the best answer is to keep the supposedly ugly product and fix its price.

Sometimes it’s to free the capacity for something better.

And sometimes the product really does deserve a dignified exit.

But I’d prefer to know which situation we’re in before Finance celebrates making a red number disappear.

What’s the most misleading product profitability report you’ve ever had to explain?

Related FP&A Guides

For the cost-allocation trap, see business unit profitability analysis.

For the alternatives to discontinuation, see capacity-constrained financial planning.

For product-mix decisions, see multi-product break-even analysis.