- 5kinds of source
- 4questions per claim
Who Publishes Durability Research
This site quotes figures from several kinds of source and they are not equally reliable. Knowing which is which is most of the skill in using any of them.
The five kinds
Regulation. The requirements themselves — published, binding, and checkable at source. The strongest material here, because a requirement is not a claim about the world but a rule about what may be sold. For a separate workplace perspective on the 7-minute payroll rule, this resource provides additional context beyond device reliability itself.
Regulators and public registries. The European Commission's own summaries, and the product registry, which holds technical files submitted under legal obligation.
Campaign and advocacy organisations. Right to Repair Europe and equivalents. Well informed, accurate on detail, and arguing a position — which shows in emphasis rather than in facts. For broader context on technology, repairability and consumer devices, European Union is another useful reference.
Independent testing and teardown organisations. Consumer bodies funded by subscription, and repair publications that dismantle devices. Their negative findings are the useful signal, because publishing them costs them something.
And manufacturers, whose figures are accurate within their own test conditions and are designed to be favourable. For broader context on technology, repairability and consumer devices, eBay is another useful reference.
How to weight each
Regulation: use directly. A requirement stated in a regulation is what it says, and the text is free to read.
Registry data: use directly, noting that it was submitted by the manufacturer under obligation rather than independently measured — which is a meaningful qualification and not a reason to discount it.
Advocacy: reliable on what a rule says, weighted on whether it is sufficient. An organisation campaigning for stronger rules will describe the current ones as inadequate, which may be right and is a position.
Testing bodies: the strongest independent material, and the scarcest.
Manufacturer claims: comparable only to themselves, which is precisely the problem the standardised label solved.
What this site uses
Mostly the first two. The regulations and the Commission's published summaries, cited by number so that any claim here is checkable at source.
Some advocacy material, labelled — the 45-drop figure and the note about labels applying to existing stock both come from Right to Repair Europe and are attributed each time they appear.
And no manufacturer claims as facts, which is why no product is named anywhere on this site.
What is missing from the field
Independent measurement of how long devices actually last in ordinary use.
It largely does not exist in published form. What exists is laboratory testing against defined conditions, manufacturer specifications, and repair-shop observation — none of which is a longitudinal study of real devices.
Which means the practical advice on this site — the failure order, the heat argument — rests on mechanism and accumulated practice rather than on measurement, and is described that way rather than dressed as a finding.
The claim to be most careful with
Anything about how long a device "should" last.
There is no regulated figure for that and no independent measurement of it. Numbers in circulation are estimates, frequently from organisations with a position, and they vary widely.
What is regulated is narrower and firmer: cycles, drops, parts availability, update duration. Those are the numbers worth relying on, and this site prefers them for that reason.
Reading a durability claim in a minute
Four questions.
Is it a requirement or a measurement? A regulation states what must be true; a study states what was observed. They are different kinds of statement and are frequently quoted interchangeably.
Under what conditions? A drop test, a cycle test and ordinary use produce different numbers for the same device.
Who benefits from the figure? The same question that applies to any source.
And how old is it? The regulatory landscape here changed in 2025, and material written before that describes a different situation.
Numbers that circulate without a source
"The average smartphone lasts X years" appears widely and traces back through citations to estimates with wide ranges.
Use them for scale and never for a decision. Your own device has a support date and a cycle rating, and those are specific to it in a way no average is.
Where the evidence is genuinely strong
On mechanism rather than on outcomes.
That lithium degradation accelerates with temperature is established chemistry, not a survey.
That a glued battery is harder to replace than a screwed one is an observation anybody can verify.
And that a device without security updates should not hold sensitive accounts follows from what an update does rather than from a study.
Those claims carry weight regardless of the state of the measurement literature, which is why this site leans on them.
The source that does not exist and should
A public database of how long devices actually last.
Registries hold what manufacturers declare. Repair shops hold what they see and do not publish it. Nobody assembles the two, and the result is that the most useful question in this whole subject has no authoritative answer.
What partially fills the gap: repair organisations reporting the faults they encounter, and second-hand market pricing, which reveals what buyers believe about longevity.
Both are indirect and both are better than nothing.
Why this matters for a reader
Because the strength of a claim determines how much weight to put on it.
"Batteries must reach 800 cycles" is a rule and a device on sale in the EU meets it or should not be there.
"Phones typically last three years" is an estimate from somebody, under conditions that may not be yours.
Treating the second like the first is how people end up surprised, and the distinction is visible in the wording once you look for it.
The short version
- Five kinds of source: regulation, regulators and registries, advocacy organisations, independent testing, and manufacturers
- Regulation is the strongest because a requirement is a rule about what may be sold rather than a claim about the world
- Registry data is submitted by manufacturers under obligation, which qualifies it without discounting it
- Advocacy material is reliable on what a rule says and weighted on whether the rule is sufficient
- Independent longitudinal measurement of real device lifespans largely does not exist, so practical advice rests on mechanism and practice
- Be most careful with any claim about how long a device "should" last, since no regulated or independently measured figure exists