“OEM” and “ODM” are casually used as synonyms in tracker sourcing — Alibaba listings, sales decks, and procurement RFPs blur them constantly. They are not the same thing. The wrong choice costs 30-50% extra in NRE, adds 8 weeks to lead time, and locks you into IP arrangements you may not want.
This article lays out the actual definitions, side-by-side cost structure, when each model fits, and what to confirm with your supplier before signing the PO. We do both at A9 — and we’ll explain why a buyer should pick deliberately rather than letting the salesperson pick for them.
Strict definitions
The distinction matters because it determines who designs the product:
-
OEM (Original Equipment Manufacturer) — the buyer brings the design (mechanical, electrical, firmware specs); the factory builds it. The buyer owns the design IP and the BOM, supplies or co-develops tooling, and pays NRE costs. The factory is a manufacturing service provider.
-
ODM (Original Design Manufacturer) — the factory brings a finished design (often an existing product line); the buyer adds branding, optional firmware tweaks, packaging. The factory owns the design IP, has already amortised the tooling, and ships from a shared platform. The buyer is buying a productised SKU with a logo on it.
Several adjacent terms describe variants of these:
- Private label — usually means ODM with the buyer’s logo
- White label — same as private label, used more in software / accessory contexts
- Contract manufacturing — usually means OEM (the factory just builds what you specify)
- Joint design / co-development — a hybrid: factory contributes a starting platform, buyer pays for custom mechanical or firmware work
If a supplier won’t tell you crisply which of these they are offering, that’s a procurement red flag — they’re either hiding which it is (probably ODM dressed as OEM) or they don’t know themselves.
Side-by-side comparison
| Dimension | OEM | ODM |
|---|---|---|
| Who designs | Buyer | Factory |
| NRE (non-recurring engineering) | USD 20,000–100,000+ | USD 0 |
| Tooling / mould fee | USD 10,000–50,000 (buyer pays) | USD 0 (amortised by factory) |
| MOQ | 500–1,000 units typical | 100–300 units typical |
| Lead time from PO | 12–16 weeks (incl. tooling) | 4–6 weeks |
| Unit cost (same BOM) | 10–15% lower at scale | Standard |
| IP ownership | Buyer | Factory (buyout possible) |
| Customisation depth | Anything | Firmware/branding/packaging only |
| VoLTE / carrier certification | Run for buyer’s specific build | Use factory’s existing certification |
| Time to market | Slowest | Fastest |
The single biggest variable is time to market. ODM gets you from PO to landed inventory in roughly 6 weeks. OEM realistically takes 4 months because you have to amortise tooling design, mould fabrication, EVT/DVT/PVT cycles, and certification (FCC / CE / regional carrier acceptance) before you can ship.
If your business model needs hardware on-shelf by a specific date, OEM is the wrong path unless you started months earlier.
When OEM wins
OEM is the right answer when any of the following is true:
- The buyer already has industrial design, mechanical engineering, and electrical design teams
- The product is genuinely differentiated from anything on the market (no existing ODM platform fits)
- Annual volume justifies amortising USD 30-100k+ NRE (typically 10,000+ units/year)
- The buyer needs to own the IP (resale brand, defensive patent positioning, licensing strategy)
- Regulatory requirements demand product-specific certification (medical, judicial, automotive)
- Long product lifecycle (3–5+ years) makes the tooling investment pay back
For trackers specifically, OEM makes sense for premium wearable brands building their own SKU at scale, B2B operators with proprietary backend integration that requires firmware customisation deeper than what an ODM platform allows, and government / judicial procurement where the buyer is required to own the design for audit reasons.
The trade-off: slow, expensive, all the engineering risk sits with the buyer. If yield is bad, BOM components go obsolete, or the certification fails, the buyer eats the loss.
When ODM wins
ODM is the right answer when all of the following are true:
- The buyer needs to go to market in under 3 months
- Annual volume is under ~10,000 units (NRE doesn’t amortise)
- The factory’s reference design meets 80%+ of the buyer’s spec
- The buyer’s value-add is brand, distribution, software, or service — not hardware design
- The buyer is fine with the factory selling similar hardware to other buyers (with different branding)
- The certification the factory already holds matches the buyer’s target markets
For trackers, ODM is the default for distributors entering a new market, system integrators bundling hardware with a software platform, and consumer brands testing a new vertical without committing to in-house hardware.
The trade-off: the buyer can’t differentiate hardware. If three distributors all buy from the same ODM factory, three nearly identical products end up in market — competition collapses to brand, price, and software.
A decision tree for procurement
Walk through these in order:
- Do you need to be in market within 90 days?
- Yes → ODM
- No → keep going
- Is your annual volume above 10,000 units?
- No → ODM (NRE doesn’t amortise below this)
- Yes → keep going
- Is your design materially different from any existing platform?
- No → ODM with custom branding/firmware
- Yes → OEM
- Do you need to own the design IP for resale, licensing, or audit reasons?
- No → ODM
- Yes → OEM, or pay for IP buyout from the ODM factory (negotiable but expensive)
- Do regulatory requirements demand product-specific certification?
- No → ODM with factory’s existing certs
- Yes → OEM
Roughly 85% of B2B tracker sourcing falls into ODM cleanly. OEM makes sense for the remaining 15% — well-funded brands at scale, regulated verticals, or true product differentiation. Many buyers start ODM to test the market, then graduate to OEM once volume justifies the investment.
What to ask suppliers before you order
Questions to ask if you’re considering ODM
- Platform age — how old is the reference design? (Modems get end-of-lifed; certifications expire.)
- Other buyers on the same platform — how many distributors are selling the same hardware? Where geographically?
- Customisation depth — what’s the maximum firmware divergence allowed without re-certifying?
- IP buyout terms — can the buyer purchase exclusive rights to the design later? At what price?
- Carrier certification list — which MNOs has the factory’s reference build passed? Are any of your target carriers missing?
Questions to ask if you’re considering OEM
- NRE breakdown — what’s the engineering cost split (mechanical design / electrical / firmware / certification)?
- Tooling ownership — who owns the moulds after fabrication? Can they be transferred to another factory?
- Yield commitment — what’s the contractual first-pass yield target? Who eats the cost if it’s missed?
- Lifecycle support — how long after end-of-production will the factory hold spare parts and component substitution rights?
- Joint vs sole IP — is the design 100% buyer-owned, or does the factory retain any joint rights?
A supplier who answers these crisply has done OEM before. A supplier who deflects (“we can discuss after PO”) is either inexperienced or hiding terms that should be negotiated upfront.
A9’s position
The A9 platform supports both:
-
ODM is our default. Standard MOQ 100 units, 25–35 days from PO to FOB Shenzhen, full firmware re-branding (logo, language, app skin, SOS routing), custom packaging and accessory kits. Carrier certifications already in place for major EU / US / LATAM bands. This is the right fit for distributors entering a new market quickly, B2B platforms bundling hardware with software, and judicial monitoring providers who need certified ankle-config hardware without designing from scratch.
-
OEM is available for partners with 1,000+ unit annual commitment. NRE quoted per project; we provide co-development on mechanical, electrical, and firmware. Tooling is buyer-owned and transferable. Carrier acceptance is run for the buyer’s specific build. This is the right fit for premium wearable brands, regulated procurement, and any deployment where hardware differentiation is the value.
If you’re scoping a tracker programme and want a no-BS conversation about which model fits your volume, timeline, and differentiation strategy, reach out — we’d rather route you to the right model upfront than oversell the wrong one.