5 ms·

How are planned obsolescense implemented in practice?

simias 9y ago
Mostly by very tight tolerances on certain components that causes them to fail prematurely. The difference between planned obsolescence and just cheapening out on the BOM is mostly in the eye of the beholder. There's also the classic "software driven obsolescence", such as pushing a software upgrade that performs poorly on older hardware. Again, you could argue that maintaining older versions of the software is costly for the vendor and it makes sense for them to push their newer, shinier version. I think in many cases "planned obsolescence" is not so much about some evil corporate type saying "let's have this thing break on purpose two years from now" but rather some greedy corporate type saying "let's make this thing as cheap as possible".
honestoHeminway 9y ago
Plastic bearings, where a metal one would last half a eternity. Capacitators near heatsources. Subconciously annoying the consumer, by making a product via software update laggy and unresponsive.
IshKebab 9y ago
It isn't. What really happens is cost reduction (https://en.wikipedia.org/wiki/Cost_reduction https://en.wikipedia.org/wiki/Cost_reduction). Basically the design is changed to make it as cheap as possible to manufacture, while still lasting the guarantee period. See https://en.wikipedia.org/wiki/Design_for_manufacturability https://en.wikipedia.org/wiki/Design_for_manufacturability Examples of what you might do: * Change the design to relax tolerances. * Switch from nuts/bolts to other fixings like glue, snap-fit, screws into plastic, etc. * Reduce the thickness of plastic parts that rarely fail * Switch to cheaper materials, e.g. metal->plastic * Integrate electronics into fewer places / components / PCBs. I have a theory that first generation products last longer than recent ones because they haven't had a chance to make them crap yet (though this may also be selection bias). None of it is really done to screw the customer over, rather to maximise profit. You can't compete on reliability nearly as easily as on price. It takes many years and lots of reviews for customers to learn the reliability of a product. It will probably be obsolete by the time its reliability is clear to the average person, so really only huge long-lived brands can compete on reliability. Cost is easy to compare.
- petra 9y ago
  You can offer some competition on reliability relatively easily. For example Asus publishizes the fact they use high quality long-lasting capacitors on motherboards. Further options: 3rd party reliability testing as branding, tear down by trustworthy experts, easy to see repairability, guarrateed access to affordable repair parts
- kuschku 9y ago
  > You can't compete on reliability nearly as easily as on price. It takes many years and lots of reviews for customers to learn the reliability of a product. It will probably be obsolete by the time its reliability is clear to the average person, so really only huge long-lived brands can compete on reliability. Cost is easy to compare. That depends. There’s a reason people buy Miele. You can also make reliability and quality your business model.
21 9y ago
It could be something like in Formula 1, where an engine is built just strong enough to hold for a weekend of usage. Because then it will be lighter, so it will have a competitive advantage.
- emn13 9y ago
  But that example is also great for another reason: very slight changes and a slightly less aggressive driving might well achieve orders of magnitude greater life expectancy at virtually no performance difference. The problem specifically for F1 is that even if you lose just 0.1% performance, that matters hugely (as in, a 0.1% speed difference seem to be a fairly normal 1st vs. 2nd spot difference, you can browse here: https://www.formula1.com/en/results.html/2017/races/959/australia/race-result.html https://www.formula1.com/en/results.html/2017/races/959/aust...)! Afterall, F1 is a competition: winner takes all (well, almost all). That last 0.1% really matters. For a fridge? Or really for almost everything people actually use? Nobody cares about that last 0.1% - heck, I wouldn't notice huge gaping performance chasms such as say 2%, but that would allow much, much better quality. Frankly, I doubt the advantages of skimping on longevity really amount to much in almost all cases.
friendzis 9y ago
To put it simply - bathtub curve [0]. Basically, every component has certain failure rate (probability that an item will fail after certain amount of use, usually hours) profile, which tends to resemble profile of a bathtub. It turns out to be possible to engineer a product in such a way, that combined probabilities form: a) very steep bathtub profile (critical/expensive components fail one after another) b) very flat bathtub profile (critical/expensive components have highly differing lifetimes) c) something in between Planned obsolence is a [0]: https://en.wikipedia.org/wiki/Bathtub_curve https://en.wikipedia.org/wiki/Bathtub_curve --- edit --- Reading other comments, I think it is important to make a distinction between terms "planned obsolescence" and "engineered to fail". My comment was (and the proposed law, as I understand, mostly is) regarding the latter, others commented on specific ways it can be implemented. Even though these terms are used in various ways, obsolete product is still technically usable (at least to some degree), but is unattractive, malfeatured or expensive to maintain, compared to current competition, often all in conjunction.

Mostly by very tight tolerances on certain components that causes them to fail prematurely. The difference between planned obsolescence and just cheapening out on the BOM is mostly in the eye of the beholder. There's also the classic "software driven obsolescence", such as pushing a software upgrade that performs poorly on older hardware. Again, you could argue that maintaining older versions of the software is costly for the vendor and it makes sense for them to push their newer, shinier version. I think in many cases "planned obsolescence" is not so much about some evil corporate type saying "let's have this thing break on purpose two years from now" but rather some greedy corporate type saying "let's make this thing as cheap as possible".

honestoHeminway 9y ago

Plastic bearings, where a metal one would last half a eternity. Capacitators near heatsources. Subconciously annoying the consumer, by making a product via software update laggy and unresponsive.

IshKebab 9y ago

It isn't. What really happens is cost reduction (https://en.wikipedia.org/wiki/Cost_reduction https://en.wikipedia.org/wiki/Cost_reduction). Basically the design is changed to make it as cheap as possible to manufacture, while still lasting the guarantee period. See https://en.wikipedia.org/wiki/Design_for_manufacturability https://en.wikipedia.org/wiki/Design_for_manufacturability Examples of what you might do: * Change the design to relax tolerances. * Switch from nuts/bolts to other fixings like glue, snap-fit, screws into plastic, etc. * Reduce the thickness of plastic parts that rarely fail * Switch to cheaper materials, e.g. metal->plastic * Integrate electronics into fewer places / components / PCBs. I have a theory that first generation products last longer than recent ones because they haven't had a chance to make them crap yet (though this may also be selection bias). None of it is really done to screw the customer over, rather to maximise profit. You can't compete on reliability nearly as easily as on price. It takes many years and lots of reviews for customers to learn the reliability of a product. It will probably be obsolete by the time its reliability is clear to the average person, so really only huge long-lived brands can compete on reliability. Cost is easy to compare.

petra 9y ago
You can offer some competition on reliability relatively easily. For example Asus publishizes the fact they use high quality long-lasting capacitors on motherboards. Further options: 3rd party reliability testing as branding, tear down by trustworthy experts, easy to see repairability, guarrateed access to affordable repair parts
kuschku 9y ago
> You can't compete on reliability nearly as easily as on price. It takes many years and lots of reviews for customers to learn the reliability of a product. It will probably be obsolete by the time its reliability is clear to the average person, so really only huge long-lived brands can compete on reliability. Cost is easy to compare. That depends. There’s a reason people buy Miele. You can also make reliability and quality your business model.

21 9y ago

It could be something like in Formula 1, where an engine is built just strong enough to hold for a weekend of usage. Because then it will be lighter, so it will have a competitive advantage.

emn13 9y ago
But that example is also great for another reason: very slight changes and a slightly less aggressive driving might well achieve orders of magnitude greater life expectancy at virtually no performance difference. The problem specifically for F1 is that even if you lose just 0.1% performance, that matters hugely (as in, a 0.1% speed difference seem to be a fairly normal 1st vs. 2nd spot difference, you can browse here: https://www.formula1.com/en/results.html/2017/races/959/australia/race-result.html https://www.formula1.com/en/results.html/2017/races/959/aust...)! Afterall, F1 is a competition: winner takes all (well, almost all). That last 0.1% really matters. For a fridge? Or really for almost everything people actually use? Nobody cares about that last 0.1% - heck, I wouldn't notice huge gaping performance chasms such as say 2%, but that would allow much, much better quality. Frankly, I doubt the advantages of skimping on longevity really amount to much in almost all cases.

friendzis 9y ago

To put it simply - bathtub curve [0]. Basically, every component has certain failure rate (probability that an item will fail after certain amount of use, usually hours) profile, which tends to resemble profile of a bathtub. It turns out to be possible to engineer a product in such a way, that combined probabilities form: a) very steep bathtub profile (critical/expensive components fail one after another) b) very flat bathtub profile (critical/expensive components have highly differing lifetimes) c) something in between Planned obsolence is a [0]: https://en.wikipedia.org/wiki/Bathtub_curve https://en.wikipedia.org/wiki/Bathtub_curve --- edit --- Reading other comments, I think it is important to make a distinction between terms "planned obsolescence" and "engineered to fail". My comment was (and the proposed law, as I understand, mostly is) regarding the latter, others commented on specific ways it can be implemented. Even though these terms are used in various ways, obsolete product is still technically usable (at least to some degree), but is unattractive, malfeatured or expensive to maintain, compared to current competition, often all in conjunction.