How do you decide whether to discontinue a low-volume product?

This guide covers a supplier deciding whether to keep, drop or replace a slow-selling industrial chemical or material product sold to business customers. It does not cover a product a regulation already bans, a consumer product line, or the notice terms of a signed supply contract, which govern first where they exist.

Which conditions decide whether to discontinue, keep or replace a low-volume product?

Four conditions decide whether to discontinue, keep or replace a low-volume product: the direct margin, the customer approval status, the customer's switching cost and how hard the decision is to reverse. Start with the direct margin. AccountingTools advises terminating a product only if its direct cost equals or exceeds its price, because a fully loaded cost moves the same overhead to the remaining products instead of removing it (AccountingTools, 2026). Direct cost here means materials, scrap, outsourced work, unique packaging, commissions and the interest on inventory held only for that product (AccountingTools, 2026).

Then record three facts for each customer that still buys the product. First, whether the product is written into the customer's approved specification, part approval or bill of materials. Second, what the customer would have to test, file or relabel to change. Third, how much of the customer's total business with you rides on it. AccountingTools lists customer need as a reason to keep an unprofitable product when the customer's other purchases are profitable enough to subsidize it (AccountingTools, 2026).

Last, check whether you already make a substitute, and whether you could restart the product after stopping it: raw materials, retained samples, the process record and the people who know it. A product that fails the margin test but has an approval and a high switching cost is a keep or substitute case, not a discontinue case.

Which options fit which low-volume product cases?

The option that fits a low-volume product case follows from approval status, switching cost and reversibility read together, not one at a time. The decision table below sets a recommended option for the common combinations. Every row assumes the product fails the direct cost test, because a product that passes it is kept by default. Fill in the left column from the facts recorded for each customer, not from memory.

The row used most often is the third: when a customer approval names the product and you make a substitute, offer the substitute and support the customer's approval instead of announcing the end of supply. A last-time buy, where the customer orders a final quantity before an end date, is the usual form of the run-out option. When two rows fit, take the one that keeps more options open, because a kept product can still be dropped later while a dropped one is slow to bring back.

Why is comparing expected outcomes alone not enough for a low-volume product decision?

Comparing expected outcomes alone is not enough for a low-volume product decision because the expected case tends to be optimistic, and the costly outcomes sit in the bad cases. UK Treasury appraisal guidance calls this optimism bias, "the proven tendency for appraisals to be over-optimistic about key assumptions", and tells practitioners to adjust for it (HM Treasury Green Book, 2026). The same guidance always takes the business-as-usual option forward to the shortlist, where "it acts as a benchmark against which other options are compared" (HM Treasury Green Book, 2026).

For a low-volume product, that means four cases for each option, each written as a difference from keeping the product as it is. The table below is an illustrative comparison for a product that one approved customer buys; the cases are examples, not data.

In this illustrative case, discontinuing looks best in the base case and worst in the bad cases, which is why a decision made on the base case alone often goes wrong.

What does a customer approval cost when a low-volume product is discontinued?

A customer approval costs the customer a new qualification when a low-volume product is discontinued, and that cost lands on the customer, not on the supplier who made the decision. In automotive supply chains, the AIAG Production Part Approval Process (PPAP) manual, 4th edition, lists the use of a material other than the one in the previously approved part, and a change of supplier for non-equivalent materials, among the changes that need customer notification (Accuride Corporation, citing PPAP Table 3.1, 2026). Ford requires its suppliers to obtain Ford approval before implementing any supplier-initiated change (Ford Customer-Specific Requirements for PPAP, June 2026). In its answer on layout inspection (clause 8.6.2), an IATF FAQ notes that product requalification "would normally imply full validation to all product approval requirements", such as a PPAP submission (IATF, June 2026).

Outside automotive, the cost takes the same shape: lab tests, a plant trial, document updates and the time of the customer's quality staff. A converter that sources replacement materials puts it plainly: "Looking at a data sheet is not enough to say whether it's similar enough to replace the original product" (Strouse, 2025). One adhesive formulator puts finding, testing and validating a replacement raw material at two to 12 months (Appli-Tec, 2020).

Ask each approved customer what its own change would take before you decide. That answer is the switching cost in the decision table, and it is a number only the customer can give you.

How hard is it to reverse a low-volume product discontinuation?

Reversing a low-volume product discontinuation is usually harder than the original decision, because what a restart needs starts to disappear when production stops. AccountingTools lists the chance to stop using certain suppliers as a downstream benefit of dropping a product with unique raw materials (AccountingTools, 2026); the same step removes the raw material source a restart would need. Retained samples age, batch records lose their owner and the operators who knew the product move to other work.

The customer side reverses badly too. The PPAP manual, 4th edition, lists product produced after the tooling has been inactive for volume production for twelve months or more as a change that needs customer notification (Accuride Corporation, citing PPAP Table 3.1, 2026). In automotive supply chains, a restart after a long pause can therefore face the approval process again.

UK Treasury guidance says decision trees can clarify alternatives "where decisions taken are either irrevocable or expensive to reverse" (HM Treasury Green Book, 2026). For a product, the practical version is a short list: what you would have to rebuy, requalify or relearn to restart, and how long each item takes. If the list is long, prefer a run-out to a hard stop, keep the formula and process record, and keep retained samples for their shelf life.

What signals should change a decision to discontinue a low-volume product?

The signals that should change a decision to discontinue a low-volume product are the values at which another option becomes better, and they should be written down before the decision is announced. UK Treasury guidance calls these switching values, "the values that a key assumption would need to change to, in order to make an option no longer value for money" (HM Treasury Green Book, 2026). Set one signal for each key assumption:

• Direct margin: the raw material cost or order volume at which the product passes the direct cost test again.

• Customer approval: a customer reports that its requalification will take longer than your run-out period.

• Substitute: the substitute fails a customer trial or needs a change the customer cannot accept.

• Restart: a supplier of a unique raw material announces its own discontinuation, which turns a reversible stop into a permanent one.

• Dependent sales: the customer ties orders for other products to this one.

Name who watches each signal and how often, and review the decision when one is crossed. A signal agreed in advance keeps the review about the evidence, not about who argued for the original choice. Do not attach probabilities you cannot support; a plain statement of what would change the decision is more useful to the people who approve it.

Common mistakes when discontinuing a low-volume product

The most common mistakes when discontinuing a low-volume product come from deciding on the supplier's own numbers alone.

1. Using a fully loaded cost. Allocated overhead stays in the business and moves to the other products, so AccountingTools advises using direct costs only (AccountingTools, 2026).

2. Skipping the before-and-after view. AccountingTools advises running an estimated before-and-after income statement and not cancelling the product if profits do not improve (AccountingTools, 2026).

3. Not asking approved customers what a change would cost them. The switching cost is the customer's figure, and a supplier estimate tends to be low.

4. Comparing the options only with each other. Keeping the product is an option, and the others should be measured against it.

5. Missing dependent sales. AccountingTools notes that dependent products can earn enough to offset losses on the main product (AccountingTools, 2026).

6. Announcing a hard stop with no run-out or last-time buy, which leaves an approved customer no time to qualify an alternative.

7. Deleting the restart path by selling off unique raw materials and retained samples before the customers have moved.

Who should review a decision to discontinue a low-volume product?

Sales, quality and production should check a decision to discontinue a low-volume product together, and every customer that holds an approval should be asked before the decision is final. Sales knows which accounts depend on the product and what else they buy. Quality knows which customer specifications, part approvals and change notification duties name the product, including customer-specific rules such as Ford's prior approval requirement (Ford Customer-Specific Requirements for PPAP, June 2026). Production knows what a restart would need.

If your site is certified to ISO 9001, run the decision under the change control of the edition on your certificate. ISO published the sixth edition, ISO 9001:2026, on 16 September 2026, and it includes clause 8.5.6, Control of changes (ISO, 2026).

Bring in the customer's quality or purchasing contact early when the product sits in an approved part or a regulated filing, because only the customer can say what its change would take. Bring in your raw material supplier when the restart question depends on a material it may stop making. Lubinpla's feasibility review works this decision in the same order: it fixes the decision, the options and the premises first, asks the user once for any premise the records do not hold, compares best, base, worst and extreme outcomes for each option including keeping the product, and returns a recommendation with the signals that would change it.