Key takeaways
- Second-sourcing is a design activity, decided at schematic and DVT — not a purchasing reaction.
- Sole-source parts are the highest-risk line items on any BOM; flag and reduce them deliberately.
- Qualify alternates before you need them — an unvalidated substitute is not a real second source.
- Lifecycle and allocation status change monthly; treat the BOM as a living risk register.
- For the few truly single-source parts, plan buffer stock and abstraction.
Shortages are a design problem first
The 2020–2023 cycle taught hardware teams a hard lesson: the most expensive component is the one you cannot buy. By the time purchasing hits an allocation wall, the cheapest fixes — picking a part with three sources, leaving an alternate footprint — are already behind you. Resilience is designed in, mostly for free, if you decide it early.
The 7 strategies
| # | Strategy | What it buys you |
|---|---|---|
| 1 | Specify pin-compatible alternates | Drop-in swap with no layout change when the primary goes short |
| 2 | Avoid sole-source parts | Removes single points of failure from the BOM at design time |
| 3 | Keep the BOM flexible | Tolerance/value ranges and DNP options let one board accept many parts |
| 4 | Validate alternates in DVT | A tested second source, not a hopeful line in a spreadsheet |
| 5 | Track lifecycle status | Early warning on NRND/EOL before it becomes an emergency |
| 6 | Buffer long-lead parts | Strategic stock bridges the gap on the few unavoidable singles |
| 7 | Abstract critical functions | A swappable module or sub-board isolates the riskiest silicon |
How to apply each
1. Specify pin-compatible alternates
For high-risk parts, identify components from different manufacturers that share the same footprint and electrical behavior, and list them as approved alternates on the BOM from day one. A pin-compatible swap needs no respin — purchasing simply buys whichever is available.
2. Avoid sole-source parts
Run a sole-source audit on the BOM: any line with one manufacturer and one distributor is a risk. Where a unique part is not essential, choose a multi-source equivalent instead. Reserve single-source parts for places where they genuinely add value.
3. Keep the BOM flexible
Specify passives by tolerance and value range rather than exact part number where the circuit allows. Add do-not-populate options and alternate footprints for parts likely to churn. Flexibility on the board multiplies your purchasing options.
4. Validate alternates in DVT
An alternate you have never built is a guess. Populate boards with the second source during Design Verification Test and confirm they pass the same suite as the primary. Only then is it a real second source you can switch to under pressure.
5. Track lifecycle status
Subscribe to manufacturer NRND/EOL notices and re-check critical parts each build. Lifecycle status is the leading indicator of the next shortage; catching it early converts a scramble into a planned redesign.
6. Buffer long-lead parts
Some parts are unavoidably single-source — a specific SoC, a unique sensor. For those, hold strategic buffer stock sized to the lead time plus a margin, so a spike in lead time does not halt the line.
7. Abstract critical functions
Where one part dominates the risk, isolate it behind a connector — a radio module, a sensor daughtercard, a power sub-board. If that silicon goes short, you redesign a small module, not the whole product.
Build resilience in, validate it in DVT
These seven move risk from purchasing back to engineering, where it is cheapest to manage. Bake alternates into the BOM during hardware development, qualify them in DVT, and fold the choices into your DFM checklist so footprints and alternates are locked before production. A resilient BOM is the difference between a substitution and a stop-ship.
Where Fundamentum fits
A resilient BOM keeps the line running; Fundamentum, our Canadian IoT platform, keeps the fleet running afterward — fleet provisioning and device identity per unit, then governed OTA and an audit trail across hardware revisions, inside a SOC 2 Type II perimeter. When a part change ships a new revision, you manage both variants from one place; it interfaces with AWS, Azure or Google Cloud only if required. See the platform →
Frequently asked questions
What is a pin-compatible alternate?
A component from a different manufacturer that shares the same footprint and electrical behavior as your primary part, so it drops in with no layout change. Listing approved pin-compatible alternates on the BOM lets purchasing buy whichever is available without a respin.
Why validate a second source in DVT?
Because an alternate you have never built is a guess, not a plan. Populating boards with the second source during Design Verification Test and confirming they pass the same suite as the primary is what makes it a real second source you can switch to under pressure.
How do I handle a part with only one source?
Some parts — a specific SoC or unique sensor — are unavoidably single-source. For those, combine strategic buffer stock sized to the lead time plus a margin with functional abstraction: isolate the part behind a connector so a shortage means redesigning a small module, not the whole product.
Isn’t second-sourcing just a purchasing job?
No — it’s a design activity. The cheapest moves (choosing multi-source parts, leaving alternate footprints, qualifying alternates in DVT) happen at schematic and verification. By the time purchasing hits an allocation wall, the inexpensive options are already behind you.
How often should I review BOM lifecycle status?
Treat the BOM as a living risk register: re-check critical parts each build and subscribe to manufacturer NRND/EOL notices. Lifecycle status is the leading indicator of the next shortage, and catching an end-of-life early converts a scramble into a planned redesign.
Related reading
Talk to an IoT engineer — free
Book a FREE 30-minute consultation with our team. No slides, no obligation — a working session on your connectivity, platform or compliance questions.
