A last time buy is the final window to order an electronic component before a manufacturer discontinues it permanently. The right quantity balances projected demand against storage cost and forecast risk, typically 12 to 36 months of consumption, adjusted for supply chain buffer and a 5 to 10% scrap allowance.
Key Takeaways
- 12 to 36 month coverage window: industry standard practice for last time buy quantities is 12 to 36 months of forward consumption, adjusted for project lifecycle and design refresh cycles.
- Scrap allowance of 5 to 10%: buyers typically add this margin above calculated need to cover rework, testing attrition, and handling losses.
- £50K+ annual cost of reactive decisions: 72% of buyers surveyed by Accuris Electronic Parts Intelligence (March 2026, N=439) report this level of cost from late or reactive component decisions.
- Rework costs up to £250K: 85% of the same survey group face rework costs at this scale when components run out unexpectedly.
- Resistor availability at 77.1%: Lytica’s May 2026 index shows resistors as the tightest category in the basket, with processors showing the largest lead time extension at +12.56%.
What Is a Last Time Buy and Why Timing Matters
A last time buy, often shortened to LTB, is the final ordering window a manufacturer gives before formally discontinuing a component. It arrives at the end of an End of Life (EOL) notification process. The manufacturer issues a notice, usually with a specific cutoff date, and after that date the part simply stops being made. No more wafers, no more assemblies, no exceptions. Distributors may hold residual stock for a while, but once that runs out, the component is gone from the market entirely.
Vishay’s SQ4532AEY-T1_GE3, a MOSFET with a last time buy date of 23 August 2026, is a clean example of how absolute this cutoff is. There’s no grace period beyond that date. Orders placed after it simply won’t be accepted against the original part number. Buyers who miss the window are pushed straight into redesign, cross-reference sourcing, or the aftermarket, all of which cost more time and money than planning ahead would have.
Timing matters because lead times don’t behave normally once an LTB window closes. During the active buy period, suppliers can usually still schedule production runs and give realistic delivery dates. The moment that window shuts, any remaining stock becomes finite and non-renewable. Lead times that were measured in weeks can stretch to months, or collapse entirely if the part simply isn’t available at any price. Buyers who assumed they could “wait and see” often find there’s nothing left to wait for.
The Nexperia export dispute in 2026 showed how fast this shift can happen even outside a formal EOL cycle. Automotive logic devices, including the 74HCT245PW, moved from being a commodity part anyone could order with confidence to an allocated item within weeks. Buyers who hadn’t secured stock found themselves competing for limited allocation almost overnight. That’s the risk with any part approaching end of life: the transition from “available” to “scarce” rarely gives much warning, and a formal LTB notice is often the clearest signal buyers get before things tighten further.
This is why last time buy decisions deserve the same rigour as a new product introduction, not a quick guess based on last year’s usage. Get the quantity wrong on the low side, and you’re back in a sourcing scramble with no manufacturer support. Get it wrong on the high side, and you’ve tied up capital in stock that may degrade, become obsolete in your own product line, or simply sit unused for years. Neither outcome is cheap. The decision deserves a proper calculation, not a round number picked under time pressure.
Calculating Demand: The Core Formula
Every last time buy calculation starts with consumption data, not forecasts. Pull your actual monthly or annual usage figures for the part across every product line it feeds into. This is your baseline. Anything built on top of guesswork rather than real consumption history is going to be wrong in one direction or the other, and wrong in a way that’s expensive to correct later.
From that baseline, apply a coverage multiplier. The industry range sits between 12 and 36 months of projected consumption, and where you land in that range depends on how long your product using the part is expected to stay in production. A consumer product with an 18 month lifecycle doesn’t need the same buffer as an industrial control system that will still be manufactured a decade from now. Match the multiplier to the real lifecycle of the end product, not a generic rule of thumb applied across your whole bill of materials.
Several risk factors should adjust that multiplier up or down. Project lifecycle is the obvious one: longer production runs need bigger buffers. Design refresh cycles matter too. If your engineering team has a redesign already planned that will drop this part, don’t buy three years of stock for a product that’s being revised in eighteen months. Customer retention plays a role as well. If the end customer for this product has a history of extending contracts or reordering beyond the original schedule, build that into the multiplier rather than treating the current forecast as fixed.
On top of the core quantity, add a scrap and rework allowance. This is not optional padding. The Accuris Electronic Parts Intelligence Survey from March 2026, covering 439 respondents, found that 85% of buyers face rework costs reaching up to £250,000 when component issues surface mid-project. A 5 to 10% allowance above your calculated need accounts for parts lost to testing, handling damage, programming failures, or taping and reeling attrition. It’s a small addition against the total order, and it’s considerably cheaper than discovering you’re short after the manufacturer has already closed the line.
Keep the formula honest. Don’t invent precision where none exists. Consumption data plus a lifecycle-adjusted multiplier of 12 to 36 months, plus a 5 to 10% scrap allowance, gives you a defensible number you can explain to finance and to engineering alike. It won’t be perfect. No forecast is. But it’s a calculation grounded in real data and a known industry range, not a figure picked because it felt safe.
How Do You Adjust Last Time Buy Quantities for Supply Chain Uncertainty?
The base formula from the previous section assumes a stable world. It isn’t one. Lead times move, availability tightens in some categories and loosens in others, and a number that looked conservative in January can look thin by June. Lytica’s May 2026 data shows exactly that kind of movement across the component basket, and it’s worth building into your LTB maths rather than treating as background noise.
Processor lead times extended by 12.56% in that dataset, the largest movement in the basket, with helium supply constraints cited as a contributing factor. Memory extended by 10.79%, and when a category shows simultaneous price and lead time extension, that’s a supply constraint signal, not just demand running hot. Resistors sat at 77.1% availability, the tightest reading in the basket. Relay and I/O components came in at 78.4%, tightened from April and moving the wrong direction. Connectors were comparatively healthy at 86.0% availability with a smaller 3.49% lead time extension.
| Category | 2026 Signal | What It Means for LTB Sizing |
|---|---|---|
| Processor | +12.56% lead time extension | Treat as highest-risk category; lean toward the upper end of the 12 to 36 month range |
| Memory | +10.79% lead time extension, price rising in parallel | Supply constrained, not just demand driven; buffer accordingly |
| Relay & I/O | 78.4% availability, tightening since April | Monitor closely; direction of travel matters as much as the current figure |
| Resistor | 77.1% availability, tightest in basket | Don’t assume passive components are low risk by default |
| Connector | 86.0% availability, +3.49% lead time | Relatively stable; standard scrap allowance likely sufficient |
None of this means buying the maximum possible quantity regardless of cost. Storage isn’t free, and capital tied up in parts sitting on a shelf for a decade is capital not doing something else. The trade-off is between re-buy risk and holding cost, and the right answer depends on who’s asking. A defence or aerospace programme with a 20 to 30 year support obligation will tolerate a much larger buffer than a consumer electronics product with an 18 month refresh cycle. Automotive and medical programmes sit somewhere between the two, usually holding stock for the length of a vehicle platform or a device’s regulatory approval period, whichever is longer. Rail programmes often mirror defence timelines given how long rolling stock stays in service.
If you’re sourcing for a constrained category, order early and order toward the upper end of your calculated range. If you’re sourcing for a comparatively healthy category like connectors, the standard scrap allowance is probably enough and you don’t need to pad the figure further. Read our guide to reading obsolescence warning signs before the EOL notice arrives if you want to catch these shifts earlier in the cycle, before the last time buy window is the only option left.
What Are the Alternatives When You Can’t Buy the Full Amount?
Sometimes the calculated figure and the available budget don’t match. Sometimes the manufacturer caps allocation before you can place the order you planned, as happened with Nexperia’s automotive logic parts during 2026 supply stress. When that happens, you need a plan that doesn’t depend on getting everything you wanted.
Lock in alternates immediately, not after the LTB window closes. Once a part is discontinued, the pool of engineers who’ve qualified a drop-in replacement shrinks fast, and so does the pool of distributors still carrying it. Start the alternate qualification process the day the EOL notice lands, in parallel with placing your last time buy order, not as a fallback if the order falls through.
Subcontracting part of the value chain can stretch a smaller stock holding further than it looks on paper. If your constraint is finished, programmed devices rather than raw die, our IC programming service lets you hold generic, unprogrammed stock and only commit it to a specific firmware revision when an order actually needs it. That turns a single LTB quantity into something flexible across multiple product variants instead of locking it to one configuration up front. Similarly, our SMD taping and reeling service can repackage bulk or loose stock into the carrier tape format your pick
