Your supplier discontinued your best-seller: the substitution playbook

The first instinct when a supplier discontinues a product is to hunt down the same item from someone else. Sometimes that's right. Roughly half the time it isn't — because "discontinued" is information about demand fit as much as supply, and your supplier ended the SKU for a reason worth weighing before you rebuild it elsewhere. The better first move is a decision: replace the product, or retire it deliberately.

This is the substitution playbook for that decision. One scope note up front, because two different events get called "discontinued": if your supplier closed entirely, every SKU went dark at once, and the supplier-went-out-of-business salvage plan owns that — including the last-buy sizing and sell-down mechanics both events share. This playbook stays inside the narrower event: one SKU reaches end-of-life while the supplier lives, and your job is the hunt for a substitute.

First: find out which kind of discontinued you're facing

"Discontinued" arrives as a subject line, a catalog footnote, or a rep's aside, and the word covers three very different situations. Pin down which one you have before spending an hour on anything else:

• Temporarily unavailable. A material delay, a retool, a backlog. This is a late-PO problem, not an end-of-life problem — the reorder math just waits.
• Discontinued, successor exists. The supplier ended your model but sells something adjacent. This is the best case: a substitute with a warm supply line.
• Discontinued, nothing behind it. The SKU is dead in their catalog and nobody is making it. The substitution hunt starts from zero.

Ask these in writing, in one email: Is the end-of-life permanent? Is there final stock, at what price, with what deadline? Which of your SKUs does this affect, and what volumes have you been buying? Do they recommend a successor in their own line? Will they keep honoring warranty and returns on the old stock? The answers decide whether you're patching a gap or shopping for a new product.

On final stock: if a last buy makes sense, size it with arithmetic, not sentiment — bridge days × velocity, with a guardrail against hoarding. That formula and its worked example live in the closure article's day-one section, and it applies unchanged to a single-SKU final buy.

Quantify the gap before you shop for a substitute

Worked example, all figures illustrative. Your pet store sells the "Alpine" no-pull dog harness — your #3 SKU by revenue. It moves 11 units/week (estimated from in-stock sales history), retails at $38, costs you $14.50 wholesale, and you're holding 74 units.

• Cover: 74 ÷ 11 ≈ 6.7 weeks. The hunt has a deadline measured in weeks, not months.
• Annual revenue at risk if the SKU just dies: 11 × $38 × 52 = $21,776 estimated.
• Annual contribution at risk: 11 × ($38 − $14.50) × 52 = $13,402 estimated in margin.

Those two numbers set the urgency budget. If the discontinued SKU were 2% of revenue, you'd quietly retire it and move on. At this scale it's worth qualifying a replacement properly — samples, spec checks, the whole discipline. The gap math is also what stops both classic overreactions: panic-buying a bad substitute to keep the listings full, or freezing and losing six weeks of a best-seller's season.

The substitution scorecard

Score every candidate against a fixed bar instead of comparing vibes. Minimums you set in advance, one row per criterion:

Three decision rules that do most of the work:

1. The 80% rule, with a safety override. A substitute hitting 80% of the spec at similar margin usually beats an empty shelf — but safety-critical attributes never get averaged. On pet gear that means buckles, stitching, and load ratings are pass/fail; a harness that fails those fails outright, whatever its margins look like.
2. Margin fail means reprice or retire. If candidates pass on spec but fail on margin, the old SKU's price was subsidized by the old cost. Either the retail price moves or the SKU retires; switching to a substitute you sell at a loss just relocates the problem.
3. Two passing candidates beat one. Take the faster lead-time candidate now and qualify the other in parallel — the structure for that is exactly what dual sourcing builds in calm times, and a discontinuation is the forced audition.

And if nothing passes: retire the SKU on purpose. The demand doesn't have to be replaced by a twin — it can flow to your closest existing product, which changes the math for the neighboring SKUs in ways worth reading about before you assume the gap closes itself (product cannibalization covers when a neighboring SKU absorbs demand and when it splits it).

Managing the crossfade from old SKU to new

The transition period — old stock selling down while the substitute ramps — is where merchants do the most avoidable damage. Four mechanics:

• New product, new SKU code. Never reuse the old code for different goods. It poisons the demand history you're about to lean on and corrupts every report that touches it. If a code genuinely must migrate, that's a deliberate migration project, not a spreadsheet find-and-replace.
• Forecast the substitute from zero history. Use the old SKU's demand as the prior, adjusted for spec differences (fit range, price point, colorway count). The zero-history forecasting methods translate directly — forecasting demand for new products covers building a first buy from an analog product's curve.
• Expect demand to split while both run. Old units and new units compete for the same buyers during the overlap, so neither looks as fast as the original did. That's the crossfade working, not failing — judge the substitute on its own curve after the old stock clears.
• Give the old SKU an exit plan with dates. Announcement to customers when you have a successor and a date; final units sold as "retiring" once the substitute is live; then archive the old listing cleanly — the retirement mechanics in Shopify (archiving, redirects, what to do with the product record) are their own end-to-end procedure, covered in retiring a discontinued product in Shopify.

One velocity note during the overlap: read the old SKU's numbers on in-stock days only. A harness that went dark for ten days mid-crossfade will look like it died of natural causes when it just ran out — and that distortion can push you into sizing the substitute's first buy wrong. (This is the same in-stock-only correction we build into InventoryX's velocity; what we can't do for you is find or score the substitute — that judgment stays yours.)

Common questions

What does it mean when a supplier discontinues a product? The SKU has reached end-of-life: no future production, with remaining stock sometimes available for a final buy at list price. It's distinct from a backorder or material delay, where production continues. Confirm in writing which situation you have before replanning anything around the product.

How do I find a replacement product my supplier no longer makes? Work spec-first, not supplier-first: write the old SKU's full spec, set minimum bars (spec match, margin, lead time, MOQ against your velocity), and score candidates against the card. Take the fastest qualified candidate, keep samples flowing on a second, and never award volume before samples pass.

Should I buy up the remaining stock when a supplier discontinues an item? Only if it's sized by arithmetic: enough to cover demand until a substitute delivers, plus buffer, capped at what you could clear if the line dies anyway. A final buy made on nostalgia becomes the dead stock you were trying to avoid — the closure-salvage article shows the sizing formula and guardrails.

When should I retire a discontinued product instead of replacing it? When safety-critical spec items can't be matched, when no candidate holds margin at a sane retail price, when the SKU is a small revenue share, or when its demand was already fading before the notice. Retirement done deliberately beats a substitute sold at a loss or a half-matched product that reviews badly.

What do I do with the last units of a discontinued SKU? Sell them down on a dated plan: a "retiring" push once the successor is live, then bundle or liquidate whatever lingers. The markdown-to-liquidation ladder and last-buy sizing are shared salvage mechanics covered in the supplier-closure playbook — reuse them; don't improvise a third version.

Next step

Don't wait for the notice: pick your top three single-sourced best-sellers and write each one's full spec sheet today, then score your current supplier against the substitution card — that baseline is what makes the future version of this decision take an afternoon instead of a quarter. The structural fix for single-source exposure, built before any discontinuation, is qualifying one backup per critical SKU.