How to Source Components Without Getting Burned

Electronics components scattered on workbench with magnifying glass examining authenticity markings

A founder recently told me about their first pilot run: 200 units of a consumer health device. They’d finalized the BOM, wired $38K to their contract manufacturer, and waited. Six weeks later, they learned that their IMU was on a 26-week lead time, a reel of 3,000 MLCCs sourced from a broker had been remarked (wrong voltage rating, discovered only after reflow), and Section 301 tariffs on their China-manufactured power management IC added 25% they hadn’t budgeted for. Total damage: four months of delay, $12K in scrap, and a pilot that shipped at negative margin.

This isn’t unusual. It’s practically a rite of passage for hardware startups in the sub-10K unit range. You have maximum exposure to supply chain risk and minimum leverage to do anything about it.

The uncomfortable truth: sourcing is not an admin task you hand off after the schematic is done. It’s a strategic activity, and the prototype-to-pilot phase is exactly where the most expensive mistakes get baked into your product, often invisibly, until it’s too late to change course cheaply.

Here’s how to avoid that.

Where Components Actually Come From (And Why It Matters)

Not all parts arrive through the same pipeline, and the channel you buy from determines your quality risk, pricing, and lead time reliability.

Authorized distributors like Digi-Key, Mouser, Arrow, Future Electronics, and Newark are your bread and butter at startup volumes. They buy directly from manufacturers, guarantee authenticity, and offer MOQs as low as one unit. The tradeoff: pricing reflects low volume. You’re paying retail.

Manufacturer direct means buying from TI, STMicro, or Murata themselves. Pricing improves substantially, but MOQs typically start at 1,000–10,000+. Rarely relevant until you’re placing pilot-scale orders for high-value ICs.

Brokers and independent distributors fill gaps when authorized channels are out of stock. This is where quality risk skyrockets. Brokers buy excess inventory, closeout stock, and sometimes factory rejects. Some are reputable. Many are not. Treat this channel as a last resort, never a default.

Your CM’s supply chain is the fourth option. Many contract manufacturers will source components on your behalf. This is convenient, but you often pay a 5–15% markup, lose visibility into where parts actually come from, and surrender BOM control. Understand what your CM is sourcing and from whom before you sign off. (More on working with CMs in [How to Choose a Contract Manufacturer].)

Rule of thumb: at sub-10K volumes, authorized distributors are your primary channel. Broker parts only with independent testing.

ChannelTypical MOQQuality AssurancePrice CompetitivenessLead Time Reliability
Authorized Distributor1–10Guaranteed authenticRetail pricingHigh (if in stock)
Manufacturer Direct1,000–10,000+Guaranteed authenticBest at volumeModerate (allocation-dependent)
Broker / IndependentVariesUnverified — buyer bewareVariable, sometimes cheaperUnpredictable
CM-SourcedVariesDepends on CM’s suppliersMarkup over direct costModerate

Counterfeits, Rejects, and the Grey Market

Counterfeit components enter the supply chain through excess inventory resellers, factory reject lots that should have been destroyed, and remarking operations that relabel cheaper parts as higher-spec versions. ERAI, the industry’s counterfeit reporting body, logs thousands of confirmed incidents per year.

The real-world impact is brutal: a single bad reel of capacitors rated at 16V instead of the marked 25V can cause field failures across your entire pilot batch. You won’t catch it during assembly. You might not catch it during functional test. You’ll catch it six weeks later when units start dying in customers’ hands.

Practical defenses:

  • Buy authorized. This is the simplest and most effective countermeasure.
  • Request Certificates of Conformance (CoCs) for any parts not sourced through authorized channels.
  • Use third-party testing for any broker-sourced parts. IDEA-standard inspection includes X-ray, decapsulation, and solderability testing. Budget $200–$500 per lot.
  • Check moisture sensitivity levels. Parts stored improperly (MSL 3 and above) may pass visual inspection but absorb enough moisture to crack during reflow soldering. Ask for dry-pack dates and humidity indicator cards.

Quality risk is the most expensive risk on your BOM. A $0.03 capacitor can generate a $15,000 rework bill.

The MOQ Reality Check

MOQs trip up first-time builders because they vary wildly by channel and aren’t always obvious until you try to order.

ChannelTypical MOQ Range
Catalog Distributor (Digi-Key, Mouser)1–10 units
Regional Distributor100–500 units
Manufacturer Direct1,000–10,000+ units
Custom / ASIC25,000+ units

Here’s the catch that surprises founders: even at “MOQ of 1” distributors, packaging minimums can force your hand. A full reel of 0402 resistors is 5,000 pieces. You can buy cut tape (smaller quantities pulled from a reel) at a per-unit premium, sometimes 3–5x the reel price. For a $0.001 resistor, nobody cares. For a $0.50 component across 40 line items, those premiums compound.

Tactical advice: During schematic review, flag any component with an MOQ above your build quantity. Design around commodity parts with low MOQs when possible. This is far cheaper than negotiating down an MOQ after the PCB layout is finalized.

Pricing Traps That Kill Your Margins

The price you pay at prototype quantities tells you almost nothing about your production cost. An STM32 microcontroller might cost $6.50 at quantity 1, $3.80 at quantity 100, $2.10 at quantity 1,000, and $1.45 at quantity 10,000. Or the curve might be much flatter. Some specialty parts barely budge with volume.

Check price breaks at 1 / 100 / 1,000 / 10,000 during every BOM review. Don’t assume a favorable curve. (For more on structuring your BOM review, see [Understanding Your BOM: A Founder’s Guide].)

Other pricing traps that catch startups:

NRE and tooling fees for custom or semi-custom components (connectors, enclosures, flex cables, custom magnetics) often run $2,000–$15,000 and are frequently omitted from initial BOM estimates. A CM might quote you a per-unit connector price without mentioning the $5,000 mold fee.

Currency and payment terms from Asian suppliers can bite. Some quote in USD but invoice in RMB at unfavorable conversion rates, or require 100% prepayment with no recourse.

Build a landed cost model early. For every line item, your true cost is: component price + shipping + tariff + handling + waste factor (budget 2–5% attrition for SMT assembly). The BOM you show investors should reflect landed cost, not Digi-Key unit price.

Tariffs and Trade Compliance: The Silent Margin Killer

Every component that crosses a border is classified by an HTS (Harmonized Tariff Schedule) code, and the tariff rate is determined by both the component type and its country of origin.

Here’s why this matters right now: US Section 301 tariffs on Chinese-manufactured goods can add 7.5–25% to your component cost. That power management IC fabricated in Shanghai? You might owe an extra $0.40 on a $1.60 part. Across 5,000 units, that’s $2,000 you didn’t budget for.

Practical steps:

  1. Ask your distributor for country-of-origin data for every line item. Most authorized distributors can provide this.
  2. Cross-reference HTS codes using the USITC database (hts.usitc.gov).
  3. Factor tariff costs into your BOM model before selecting parts, not after you’ve committed to a supplier.
  4. Build a sensitivity range, not a single number. Tariffs can change with 30 days’ notice. Model your BOM at current rates and at +10%.

Lead Times and the Allocation Trap

The lead time listed on a distributor’s website is the “standard” lead time, meaning what the manufacturer quotes under normal conditions. Actual lead times depend on current stock, factory capacity, and your position in the allocation queue.

During the 2020–2023 semiconductor shortage, lead times for common parts (STM32 MCUs, TI power regulators, basic MOSFETs) ballooned from 8 weeks to 52+ weeks. Some parts hit 80 weeks. The shortage is largely resolved, but the lesson stands: lead times are not static numbers.

The allocation trap is especially cruel for startups. When capacity is constrained, large OEMs placing orders for 500K units get allocation priority. Your order for 2,000 pieces goes to the back of the line, or gets pushed out entirely.

Strategies to protect yourself:

  • Check live stock on Octopart or FindChips before finalizing your BOM. If a part shows zero inventory across all distributors, that’s a red flag.
  • Place early orders on long-lead-time parts. Don’t wait until your CM needs them.
  • Maintain a watch list for critical components and monitor stock levels monthly.
  • Watch for last-time-buy (LTB) notifications. If a part goes end-of-life, you may have a 90-day window to stock up, or you’re redesigning.

Add a 30% buffer to every lead time in your project plan. If the distributor says 12 weeks, plan for 16.

Single-Source Dependencies: Your Biggest Hidden Risk

Single-source means only one manufacturer in the world produces the part, not just one distributor stocking it. If Bosch is the only company making a particular IMU, or a specific Qorvo RF front-end has no pin-compatible alternative, your entire production timeline is hostage to that manufacturer’s capacity, pricing decisions, and product lifecycle.

At the startup stage, this is uniquely dangerous. You have no leverage to secure allocation, no strategic account relationship, and no fallback plan.

Common culprits: specialty MEMS sensors, specific MCU variants with unique peripherals, niche RF components, and proprietary connector systems.

Mitigation starts during schematic design, not during sourcing. (See [Design for Manufacturability Checklist] for integrating this into your design review.)

  • Flag any component with only one manufacturer in your BOM.
  • Identify a qualified alternate, ideally footprint-compatible, or design in flexibility (e.g., dual-footprint pads, software-configurable interfaces).
  • If you absolutely must use a single-source part, acknowledge the risk explicitly in your project plan, stock 6–12 months of buffer inventory, and have a redesign path sketched out.

Your Sourcing Checklist for Sub-10K Builds

Before you commit capital to a pilot run, walk through every line on your BOM against this list:

  • ☐ Run full BOM through a distributor availability check (Octopart is free)
  • ☐ Flag any single-source components and document alternates
  • ☐ Identify MOQs for every line item and verify packaging minimums (reel vs. cut tape)
  • ☐ Get firm quotes (not indicative) for your actual pilot quantity
  • ☐ Confirm country of origin and applicable tariff rates for every component
  • ☐ Check lead times against your build schedule and add a 30% buffer
  • ☐ Decide buy-direct vs. CM-sourced for each line item and document who owns procurement risk
  • ☐ Document qualified alternates for all critical components
  • ☐ Calculate landed cost (component + shipping + tariff + handling + attrition), not just unit price

Treat Sourcing as a Design Decision

The components you choose during prototyping follow you into production. A part that’s easy to hand-solder onto ten boards might be single-source, tariff-exposed, and on a 30-week lead time. You won’t discover any of that until you’re trying to build 2,000 units on a deadline.

Build sourcing awareness into your BOM reviews from day one. Every component selection is simultaneously a design decision, a supply chain decision, and a financial decision. The founders who internalize this don’t just avoid getting burned. They ship on schedule with margins intact.

Ready to take the next step? [Planning Your First Pilot Run] walks through how to turn a sourcing-ready BOM into an actual production plan.


Hubble Network connects your devices via Bluetooth to satellite—no ground infrastructure, no carrier contracts. See how it works →