Last-Time-Buy Decisions: Stockpile, Bridge Buy, or Redesign?
An EOL notice with a last-time-buy deadline forces one of three choices: lifetime stockpile, bridge buy, or redesign. This framework covers the demand math, the hidden costs, and the decision criteria for each.
When an EOL notice lands with a last-time-buy (LTB) deadline, a BOM owner has exactly three options: buy enough parts to cover the product for the rest of its life, buy a smaller bridge quantity while qualifying a replacement, or commit to a redesign and buy only what carries production to the changeover. Everything else is a variation of these three.
The single question that separates them is this: how long must this design keep shipping, and can a qualified alternate realistically be ready before purchased stock runs out? If the product is near the end of its own life, a stockpile is usually cheapest. If a credible alternate exists and qualification is a matter of months, a bridge buy keeps cash free. If the part is one of several aging components on the same board, redesign often wins on total cost even though it looks most expensive up front.
This article gives you the inputs to gather, the honest cost picture for each option, a worked sizing example, and a decision table. For what triggers this decision and how to read the notice itself, see NRND vs EOL vs Obsolete: What BOM Owners Should Do at Each Stage.
Quick Answer: How to Choose
Choose a lifetime stockpile when remaining product life is short or fixed (roughly one to three years), demand is predictable, and no redesign budget exists. Choose a bridge buy when a realistic alternate has been identified and the qualification timeline is shorter than the stock the bridge quantity provides. Choose redesign when product life is long, demand is uncertain, the LTB cash requirement is large, or the board carries multiple at-risk parts that would each need their own stockpile.
All three options require the same demand inputs, so gather the data before debating the strategy.
Inputs You Need Before Any Math
An LTB quantity is only as good as its inputs. Before sizing anything, assemble:
- Remaining product life. Not the current forecast horizon — the commercial decision about how long this design will ship, including contractual supply commitments to customers.
- Annual usage. Units per assembly multiplied by realistic annual build volume, using sales input rather than last year’s actuals alone.
- Service and RMA demand. Warranty repairs, spares contracts, and field-failure replacement over the full support period, which often extends years past the last production build.
- Attrition and scrap. Assembly yield loss, test fallout, and handling damage. Small percentages compound across years of builds.
- MOQ and packaging. The manufacturer’s minimum order and reel or tray quantities set the granularity of what you can actually buy.
- Storage constraints. Moisture sensitivity level, oxidation risk on leads and finishes, and any customer date-code acceptance policies that could make old stock unusable on paper even when it is electrically fine.
The storage point deserves emphasis because it silently caps the stockpile option. Moisture-sensitive devices stored outside proper dry conditions require rebaking per JEDEC J-STD-033 handling practices before reflow, and some customers refuse date codes older than two or three years regardless of storage quality. A five-year stockpile plan that ignores date-code policy is a plan to scrap parts.
Option 1: Lifetime Stockpile
A lifetime stockpile buys the full remaining-life demand in one order before the last order date. It wins when the product’s end is visible and the total quantity is affordable: end-of-life industrial products with support contracts, low-volume equipment with fixed service obligations, or any case where redesign cost clearly exceeds the buy.
The unit price on the purchase order is not the cost of this option. Add carrying cost (capital tied up for years), storage under controlled humidity for moisture-sensitive parts, insurance, the write-off risk if the product ends earlier than planned or a mid-life redesign happens anyway, and the scrap risk from date-code policies. A common rule of thumb puts annual carrying cost at a meaningful percentage of inventory value; over a five-year horizon this materially changes the comparison against redesign.
Structure the buy to manage its own risks: split storage across locations if quantity justifies it, keep full lot and date-code documentation with the stock, and book a revisit every six to twelve months to compare actual consumption against the plan while some corrective options still exist.
Option 2: Bridge Buy
A bridge buy purchases enough stock to cover production only until a qualified alternate takes over. The sizing anchor is the alternate qualification timeline, not the product life: if qualification realistically takes nine months, the bridge is nine months of demand plus a slip allowance, not three years of it.
This is usually the most capital-efficient option, but it carries schedule risk: the plan fails if qualification slips past the stock. Protect it three ways. First, add explicit buffer — a bridge sized at qualification-time-plus-fifty-percent is common where the alternate is not yet proven. Second, define a checkpoint date on which qualification progress is reviewed while the LTB window may still be open, preserving the ability to convert to a larger buy. Third, have a fallback source identified for the original part (authorized residual stock, authenticated open market) in case both the schedule and the window are missed.
A bridge buy is only as credible as the qualification plan behind it. Footprint, electrical, behavioral, thermal, and compliance verification of the candidate part is its own workflow, covered in our guide to qualifying a pin-compatible alternate without a board respin.
Option 3: Redesign
Redesign replaces the at-risk part with an active one, accepting layout, firmware, and requalification cost in exchange for eliminating the obsolescence exposure entirely. Three conditions reliably tip the decision toward redesign:
- Multiple at-risk parts on the same board. If three components each need their own stockpile, one redesign amortizes across all of them, and mature boards rarely lose only one part at a time.
- A compliance or platform refresh is due anyway. If the product needs recertification, a cost-down, or a component-standard update in the next year or two, folding the replacement into that work makes its marginal cost small.
- No true drop-in exists. If every candidate alternate requires a layout change regardless, the “cheap” options are gone and the only question is when to do the redesign, not whether.
Redesign still needs parts until the new revision ships, so it is normally paired with a small bridge buy sized to the redesign schedule. The same slip-buffer logic applies, with the added protection that a redesign schedule is under your own control in a way a supplier qualification is not.
Worked Example: Sizing an LTB Quantity
A product ships 4,000 units per year, uses two of the affected parts per assembly, and must ship for three more years with a five-year service obligation. Assembly scrap and test fallout run 3%, and service demand is estimated at 2% of the installed base per year.
| Component of demand | Basis | Quantity |
|---|---|---|
| Production demand | 4,000 units × 2 per assembly × 3 years | 24,000 |
| Scrap and attrition | 3% of production demand | 720 |
| Service reserve | 2% of installed base per year × 5-year support × 2 per repair | ~2,400 |
| Uncertainty buffer | 10–15% on the subtotal, per demand confidence | 2,700–4,100 |
| LTB quantity range | rounded up to MOQ / full reels | ~30,000–31,500 |
Two observations generalize from this example. The service reserve is the number teams most often forget, and it is the one that cannot be corrected later — once the last order date passes, field support runs on whatever was bought. And the buffer percentage should reflect demand confidence honestly: a contract-backed forecast deserves a smaller buffer than a sales projection.
Decision Table
| Remaining product life | Credible alternate exists | No credible alternate |
|---|---|---|
| Under ~2 years | Stockpile (qualification not worth it for the tail) | Stockpile |
| 2–5 years | Bridge buy + qualify the alternate | Stockpile now, scope redesign in parallel |
| Over ~5 years or open-ended | Bridge buy + qualify; consider redesign if multiple parts at risk | Redesign + bridge buy to changeover |
Treat the boundaries as prompts rather than rules: cash constraints, customer approval requirements, and how many other parts on the board are aging all shift the lines. Lead-time and price signals on the candidate alternates matter too — an alternate that is itself on allocation is not a credible bridge target, a dynamic covered in our guide to component lead times, pricing, and supply risk.
Risk Controls After the Decision
Whichever option wins, three controls keep it honest. Document the decision — the demand inputs, the option chosen, and the revisit triggers — so that when consumption diverges from plan, the original assumptions are visible instead of folklore. Schedule revisits every six to twelve months comparing stock burn against forecast; a stockpile that is depleting fast is an early redesign warning, and one depleting slowly is a write-off warning. And keep the stock’s paperwork production-grade: lot numbers, date codes, storage conditions, and bake history for moisture-sensitive parts, because an undocumented stockpile ages into open-market stock sitting in your own warehouse.
Key Takeaways
- Every LTB decision reduces to stockpile, bridge buy, or redesign; the deciding question is remaining product life versus realistic alternate readiness.
- Gather demand inputs first — production, scrap, service reserve, MOQ, storage limits — because all three options price off the same numbers.
- A stockpile’s real cost includes carrying cost, storage, date-code policy risk, and write-off exposure, not just the purchase order.
- Size a bridge buy to the alternate qualification timeline plus slip buffer, never to product life.
- Redesign wins when multiple parts are at risk, a refresh is due anyway, or no drop-in exists — and it still needs a bridge buy to the changeover.
- Revisit the decision every six to twelve months against actual consumption while corrective options remain open.
FAQ
How do I calculate a last-time-buy quantity?
Sum production demand (annual volume × units per assembly × remaining years), scrap and attrition (typically a few percent), and a service reserve covering the full support period, then add an uncertainty buffer of roughly 10–15% and round up to MOQ or full-reel quantities. The service reserve is the component most often underestimated.
How long can electronic components be stored?
Electrically, many components store well for years under controlled temperature and humidity, with moisture-sensitive devices handled per JEDEC J-STD-033 (dry storage and rebaking before reflow). The practical limits are usually solderability of lead finishes and customer date-code acceptance policies, which can reject stock two to three years old regardless of condition.
What if the last-time-buy window is too short to decide?
Place a defensible interim order before the last order date — sized to a conservative bridge scenario — rather than missing the window entirely. It is usually possible to buy slightly more than a bridge requires; it is never possible to order after the window closes. Then finish the analysis and adjust strategy around the stock you secured.
Can I combine a bridge buy with a redesign?
Yes, and for long-life products it is the standard pattern: a bridge buy sized to the redesign-and-changeover schedule, with buffer for slips. The bridge protects production; the redesign removes the exposure permanently. The key is sizing the bridge to a schedule someone is accountable for.
