A buyer is offered a cheaper alternative to a component: same material, same nominal dimensions, same function, a reputable brand with its own certification and a shorter lead time. The switch takes a few minutes to approve, falls well within the buyer’s authority and is recorded as a saving. By any normal standard, it is a sensible decision.
It may also have removed the business’s ability to prove that the finished product meets a performance requirement on which approval, insurance or a customer warranty depends. And nothing in the normal purchasing process will flag it.
The reason lies in how high-consequence performance is usually demonstrated. Fire resistance, structural capacity, electrical safety, weather sealing, impact resistance and many similar properties are not proved for a material or part in isolation. They are proved by testing or verifying a complete assembly, built in a specific configuration, under defined conditions. The result describes that assembly: its components, arrangement, fixings and interfaces, taken together. It does not travel with any single component into a different combination. This article explains why substitutions matter, why “equivalent” is not an evidential term and how to control changes so your evidence keeps matching what you actually build.
Evidence belongs to the configuration
When a system is tested, the test report or certificate describes the specimen that was tested: which components, from which manufacturers, in what arrangement, with what fixings, sealants, gaps and spacings. Every element of that configuration is part of what the evidence covers, including the ones nobody thinks of as critical: the screw type, the sealant, the gasket, the backing material, the fixing spacing.
Change one element and you may no longer hold evidence for what you have built. You hold evidence for something similar. Whether the change matters depends on the specific requirement, the evidence and the rules that apply, which is exactly why it needs to be assessed rather than assumed.
This applies across many industries. Fire-rated building systems, electrical switchboards and control panels, machine guarding and safety systems, pressure equipment, structural connections, medical device combinations and vehicle components are all areas where compliance commonly depends on a verified combination rather than on individual parts. The specific rules differ between standards, regulators and customers, so always check the requirements that apply to your product.
Why “like-for-like” is not an evidence test
In purchasing, like-for-like means functionally and dimensionally interchangeable. That is a perfectly sound idea for procurement. But functional equivalence says nothing about evidential equivalence: whether the evidence you hold still covers the new combination. The phrase can quietly move a decision from “needs technical review” to “routine purchase” without any analysis, which is what makes it risky.
Three related misconceptions are worth naming:
- “Better cannot be worse.” A substitute that is stronger, thicker or denser in every measurable way still creates an untested combination. Assemblies under extreme conditions, such as fire, overload or vibration, do not always respond intuitively. A stiffer part can change how an assembly moves, seals or fails.
- “Change control already covers this.” Often it does not. Change control typically triggers on design changes, while a substitution is classed as a purchasing event and receives a price review, not a technical review.
- “The supplier would tell us.” Manufacturers sometimes change a product’s formulation or construction without changing its part number. Unless your supply agreement requires notification, you may never know.
Authority does not match consequence
In most businesses, purchasing authority is set by the price of the item. The consequence of a substitution, however, depends on the value of the performance claim the assembly supports, and beyond that on the value of whatever the assembly protects: a building, a machine, a person. A $5 fastener can sit inside an assembly whose failure would be catastrophic.
That mismatch is the core of the problem. People with no reason to think a decision is consequential are authorised to make it, because their authority is calibrated to price rather than to risk.
Where it goes wrong
The configuration that drifts
A product made repeatedly over years is especially vulnerable. The first version is fully evidenced. Over time, a gasket supplier reformulates, a cable gland becomes obsolete and is replaced, a sealant is switched to a preferred brand and a fixing pattern is adjusted to suit a new jig. Each change is small and individually approved. Cumulatively, the product being made is no longer the product that was tested, and no single decision was large enough to attract attention.
The assembly completed by someone else
Sometimes the final combination is assembled by a customer, installer or end user from parts supplied separately. If your evidence covers a particular combination, but the customer combines your component with another manufacturer’s parts, the result may be outside your evidence, even though each part is individually certified. If you supply components that others combine, tell customers which combinations your evidence covers.
Evidence that never matched
Sometimes the mismatch starts before any substitution. A manufacturer tests a configuration designed to show off its own product. A project then adopts that test report for a build that merely resembles it. The evidence and the installed assembly were never the same thing.
A substitution gate
The practical control is a substitution gate: a checkpoint every proposed component change passes through before it is committed. Its key principle is that the authority to approve a change is set by the consequence of the performance claim, not by the price of the part.
When the gate applies
The gate should trigger on any change to a component’s manufacturer, model, formulation, dimensions, finish, fixing method or position within an assembly that supports a performance or compliance claim. It should also trigger when a supplier notifies a change to their product, even without a part number change.
Three questions
| Question | How to answer | If the answer is no | Who can approve |
|---|---|---|---|
| Does a performance or compliance claim depend on this assembly? | Trace from the claim to the assembly, not from the part | Release as a normal purchase | Buyer |
| Is the proposed part named in the evidence we hold? | Read the test report or certificate itself, not a datasheet | Go to the next question | Technical reviewer |
| Do we hold evidence that covers the new combination? | Evidence must be in hand, not “available on request” | Refuse, or obtain new evidence before changing | Named responsible engineer |
Four rules that make it work
- No routine purchasing approval inside high-consequence assemblies. Where the claim involves personal safety, structural adequacy or a regulated performance requirement, buyers cannot approve substitutions at any price.
- Evidence in hand. A configuration is covered only if you hold a document that names it. Evidence a supplier says exists does not count until you have it.
- Evidence stays with the product or asset. Keep test reports, certificates and configuration records with the product file or asset records, so they survive handover, staff changes and sale.
- Look backwards as well as forwards. Check assemblies already built or installed within warranty or liability periods, to understand any existing exposure.
A worked example
This is an illustration. A small business builds electrical control panels for industrial customers. Its standard panel design has been verified against the requirements the customer specifies, using particular circuit breakers, enclosures, busbars and cable sizes. A customer asks for a lower price, and the buyer finds circuit breakers from a different manufacturer that are cheaper and quicker to obtain, with the same ratings.
Without a gate. The buyer approves the swap as like-for-like. Panels are built with the new breakers. A year later, during a site audit, the customer’s engineer asks for the verification records for the installed panels. The records name the original breakers. The business cannot show that the new combination has been verified, for example for temperature rise inside the enclosure, which depends on the heat produced by each component. The panels have to be assessed after the event, at the business’s cost, with the customer’s trust shaken.
With a gate. The buyer raises the substitution. The technical reviewer traces the panel to the verification evidence and sees that it names the original breakers. The responsible engineer checks the requirements and the manufacturers’ data, and decides that verification must be updated before the switch. The work takes a few days and some engineering cost. The business then offers the customer the lower-cost panel with updated evidence, and records the new configuration so future builds match it.
The saving from the cheaper breakers is real, but only after the evidence has caught up.
Build a simple configuration register
To answer the second gate question quickly, keep a register that links each performance or compliance claim to the assembly it describes and the components named in the evidence. A spreadsheet is enough for most small businesses.
| Claim | Assembly | Evidence document | Components named | Critical items | Last reviewed |
|---|---|---|---|---|---|
| Fire resistance level for a wall penetration | Pipe penetration through plasterboard wall | Test report reference | Pipe type, collar model, sealant, board type | Collar and sealant | Date |
| Enclosure temperature rise within limits | Standard control panel | Verification record | Enclosure, breakers, busbars, cable sizes | Breakers and enclosure | Date |
| Guard interlock safety function | Machine guard door | Design and validation records | Interlock switch, safety relay, wiring | Switch and relay | Date |
With the register in place, a buyer or supervisor can see immediately whether a component is critical, and the technical reviewer can answer the evidence question in minutes rather than days.
Signals worth watching
- Approval records that say “equivalent” or “like-for-like” with no evidence reference.
- Certificates held for a configuration that differs from what is built or installed.
- Supplier change notices filed without assessment.
- Cost-saving exercises that report savings on assemblies carrying safety or regulated claims.
- Handover documents that consist of brochures rather than documents naming the actual build.
How this applies to a small Australian business
Small manufacturers, fabricators, installers and contractors often work with tested or verified systems: fire-rated doors and penetrations, structural fixings, switchboards, machine guards, pressure fittings, waterproofing systems, vehicle components and more. Practical steps:
- Identify your high-consequence assemblies: those whose failure could cause injury, damage or regulatory breach.
- Keep the evidence for each: test reports, certificates and configuration records, naming the exact components.
- Mark critical components on drawings, bills of materials and purchasing records so buyers know they need review.
- Require suppliers to notify changes to products you use in critical assemblies, and write that into supply agreements.
- Train buyers and supervisors that “equivalent” needs a technical check in these assemblies.
- Tell customers and installers which combinations your evidence covers.
- Check the rules that apply: building codes, product standards, electrical regulations, customer specifications and insurance conditions. Requirements vary, so confirm with the relevant authority, certifier or a qualified engineer.
The article on proving a process is ready for production explains how to keep production evidence current when processes change, and the article on when expert advice becomes a liability covers documenting the basis of technical decisions.
When substitution is unavoidable
Sometimes you have no choice: a component is discontinued, a supplier closes or lead times become impossible. Plan for these situations rather than reacting under pressure:
- Identify single-source critical components in advance and look for alternatives before you need them.
- Pre-qualify alternatives by obtaining evidence for the alternative combination while there is time.
- Hold strategic stock of critical items with long or uncertain lead times.
- Agree interim arrangements with customers where a substitution needs evidence that cannot be produced immediately, rather than proceeding quietly.
Common mistakes
- Treating “like-for-like” as proof that evidence still applies.
- Setting approval authority by price rather than consequence.
- Classifying substitutions as purchasing events, so they bypass technical review.
- Relying on datasheets instead of reading the actual test report or certificate.
- Accepting evidence “available on request” that you have never seen.
- Ignoring silent reformulations by suppliers.
- Losing evidence when projects close or staff leave.
- Assuming each certified part makes a certified whole.
Questions to ask
- Which of our products or installations depend on tested or verified assemblies?
- Do we hold documents that name the configuration we actually build, and has anyone read them recently?
- At what authority level can a component inside those assemblies currently be changed?
- Which suppliers are required to tell us about changes to their products?
- How many substitutions were made last year, and how many received a technical review?
- If a customer, regulator or insurer asked for our evidence tomorrow, how long would it take to produce?
- Do our customers know which combinations our evidence covers?
Bringing it together
Performance evidence describes the assembly that was tested or verified, not each component within it. That makes every substitution inside a high-consequence assembly a question about evidence, even when the new part is equivalent or better. Because purchasing authority is usually set by price while the consequence depends on the performance claim, these decisions often happen where nobody checks. A simple substitution gate fixes this: trigger it on any change to a critical assembly, trace the claim to the evidence, read the actual documents and set approval authority by consequence. Keep evidence with the product, require suppliers to notify changes and tell customers which combinations are covered. Substitutions are a normal part of business. Unchecked ones in critical assemblies are not.
Source: KEVOS notes. Examples in this article are illustrations. This article is general information, not legal or engineering advice. Confirm the requirements that apply to your product with the relevant authority, certifier or a qualified engineer.
