Short answer: opening a product answers one class of question very well — how it is built — and leaves five data sets completely untouched: the exact model and production date, part numbers and supplier identity, cell capacity test results, housing weights and resin grades, and the commercial inputs. A teardown report describes what you are holding. A bill of materials is a claim about what will arrive in production, and no teardown can support that claim on its own.
What a teardown is genuinely good at
It is worth being precise here, because teardown work is not a weak method — it is a specific one. Opening a few samples and recording what you find will reliably tell you:
- How the subsystems are integrated. Where the board sits relative to the motor, how the battery is retained, whether the charging contact is a spring pin or a plated pad.
- Which parts are present and which are absent. A protection device that is not on the board is a straightforward finding, and so is a thermal path that has clearly been designed in.
- How it is assembled. Adhesive selection and where it has been applied, connector choice, wire routing, and whether anything mechanically retains a cable.
- Whether two samples are built the same way. Open two units from the same lot and compare them. Differences between the two are the cheapest early warning you can get about process control.
That is a screening method, and screening is what it should be used for — before you commit volume, and before tooling is cut. The trouble starts when the same report is used to answer questions it structurally cannot answer.
Where teardown evidence stops
Four boundaries are worth stating plainly, because each one is regularly crossed in supplier conversations:
- Cost. A teardown gives you a parts list at best. Unit cost depends on order volume, tooling amortisation, yield, labour rates and overhead allocation — none of which is a property of the object on your bench. A cost figure derived from disassembly is an estimate of material content, and it should never be presented as a landed cost.
- Specification. Two cells of the same external format can differ in rated capacity, discharge capability, protection circuit and cycle life. Two boards with the same outline can carry different firmware. The housing does not tell you which of them you have.
- Supplier identity. A marking on a component is often the assembler's or the module maker's, not the maker of the cell inside it. And even when the marking is the component maker, grade and bin within a family are not visible at all.
- Durability. One unit on a bench, observed for an afternoon, tells you nothing about a life distribution. Reliability is a property of a sample population under defined conditions, not of a specimen.
Component count is not quality
This is the most common misreading that comes out of a teardown photograph. Two boards from the same product category will differ in component count for a completely innocent reason: one of them implements more features. A higher count is a statement about function, not about quality.
What actually predicts long-term stability on a board is a shorter and more boring list, and it is worth checking explicitly:
- Protection where the energy enters. What sits between the connector and the rest of the circuit, and how close it is to the point of entry.
- Thermal management on the parts that dissipate. Copper area, thermal vias, spacing, and whether a hot component sits next to a temperature-sensitive one.
- Control logic. Whether the firmware reduces duty cycle or current as a limit is approached, rather than continuing until a protection device trips.
- Mechanical reinforcement. Adhesive or potting at joints that see vibration, and retention features at connectors and cable exits.
- Consistency between the samples you opened. The same reinforcement applied in the same places — not applied once as a fix on a single unit.
Notice that four of those five are process questions rather than design questions. That is exactly why a teardown is a poor place to settle a quality argument on its own.
The five data sets a BOM needs that a teardown cannot supply
This is the practical list we work from. Every item on it is something a supplier can provide in writing, and something that can then be verified on samples — which is the real difference between a teardown report and a specification you can hold a supplier to.
| Data set | Why the teardown cannot supply it | How it is obtained and verified |
|---|---|---|
| Exact model and production date | Two revisions can share a housing, and the unit you opened is simply whatever was shipped to you | Revision and date code locked in the purchase order, then recorded on the first-article inspection sheet |
| Part numbers and supplier identity | Markings frequently identify the assembler rather than the component maker; grade within a family is not visible | Approved vendor list and component part numbers declared, then checked for identity and grade at incoming inspection |
| Cell capacity and discharge test results | Capacity cannot be seen, and a label is a claim rather than a measurement | Discharge test at a defined rate and cut-off voltage, at a recorded temperature, plus a cycle-life test on a smaller sample |
| Housing weights and resin grades | Weight can be estimated on a scale, but the grade cannot be seen at all — and grade governs ageing and drop behaviour | Resin grade declared on the drawing, with a per-part weight check against the drawing at first article and in production |
| Packaging, labour, tooling and overhead | These are commercial inputs. None of them exists anywhere inside the product you opened | Provided as quotation line items and an amortisation schedule — which is precisely why a teardown-based cost figure is a material estimate and nothing more |
A supplier who can produce all five is a supplier you can write a specification against. A supplier who can produce two is a supplier you are currently trusting on the other three.
The battery is where the gap is widest
Cell format is one of the few things a teardown genuinely does show, and it is worth understanding what it means. Cylindrical cells and pouch cells are both standard in personal care devices, and they behave differently. A pouch cell can be shaped to fit the available space and packs more energy into a given volume, but it needs more mechanical support and more attention to swelling and puncture. A cylindrical cell brings a rigid can and a standardised protection package. Neither architecture is a quality verdict — they are two different engineering trade-offs.
What the teardown cannot tell you in either case: the delivered capacity, how the cell behaves at the discharge rate the product actually draws, whether the protection circuit is present and functional, who made the cell, and what happens to capacity after a few hundred cycles. Those five questions are the entire basis for comparing two batteries, and disassembly answers none of them.
Our rule is blunt: a capacity label is a claim until it has been measured. The label is normally a cell datasheet figure taken at a favourable discharge rate. The number a customer experiences is what the pack delivers at the rate the product draws. Those are two different measurements, and they are frequently far apart — which is why we test rather than read.
What we require before tooling is cut
In practice, the five data sets above become four gates in our own programme:
- Declared configuration. Revision, date code, approved vendor list and component part numbers, recorded in writing before sampling begins.
- First-article inspection against the drawing. Including per-part weight and declared resin grade, checked on the actual first article rather than on the drawing alone.
- Measured performance on samples. Cell discharge and capacity, plus the reliability gates that apply to that product family.
- Process-control evidence. The same reinforcement and the same routing on consecutive units — because a design that only exists on the golden sample is not a design.
The reason to insist on all four before tooling is simple. After the tool is cut, every one of them becomes expensive to change. Before it is cut, they are all just questions.
The model used as the reference in this article is the K100P vibrosonic electric toothbrush.
Frequently asked questions
Can I estimate a product's cost from a teardown?
You can estimate material content. You cannot get to a landed cost, because tooling amortisation, yield, labour rates and overhead are not properties of the object you opened. Treat a teardown-based figure as a material estimate and label it as one — the phrase to avoid is "this is what it costs to make".
Is a teardown still worth doing if it cannot give me a BOM?
Yes. It is the cheapest way to find out what class of design you are buying, and it is the only practical screening method before tooling. The failure mode is not doing the teardown — it is presenting a teardown report as if it were a specification.
How many samples should I open?
At least two units from the same lot, so you can compare them against each other. A single sample cannot distinguish a design decision from a one-off build. If the two differ in reinforcement, routing or adhesive placement, that finding matters more than anything either unit shows on its own.
What should I ask for instead of a teardown report?
Ask for the five data sets in the table above, in writing: revision and date code, part numbers and supplier identity, cell test results, housing weights and resin grades, and the commercial inputs as quotation line items. Then verify the ones that can be verified on samples.
How do I check a battery capacity claim?
Discharge the pack at the rate the product actually draws, down to a defined cut-off voltage, at a recorded temperature, and compare the result with the label. A label figure quoted at a favourable discharge rate is not a lie, but it is not the number your customer experiences either.
If you are sourcing a personal care device and want the specification side of this list rather than the teardown side, our OEM/ODM programme covers how each item is verified before volume. For the screening half of the process, see what a structured teardown audit actually records, and for an example of what a written specification looks like in practice, the 23 measurable requirements we place on a supplier. Request a quotation with your target market, volume and verification requirements.
Related reading: