[Gllug] Linux invalidates your HP warranty

Christopher Hunter chrisehunter at blueyonder.co.uk
Wed Apr 4 04:58:50 UTC 2007


On Tue, 2007-04-03 at 23:08 +0100, Bernard Peek wrote:

> HP seems to have cocked up their customer service here, possibly by not 
> explaining what they wanted.

"Customer Service" from HP (and most other vendors) is useless in most
cases - the only successful outcome I ever had from HP was getting the
address for Warranty Returns (which differed from that published on
their website!).

> What they want is to be able to do some triage, a quick and dirty fix to 
> most problems. The usual first step is to restore the machine back to 
> its as-delivered state. They could do that by swapping in a new 
> hard-disk, and if you take out the hard-disk before returning it then 
> they will have to.

It is usually not cost effective for the manufacturer to actually bother
to try to fix a broken machine.  They won't employ trained staff in this
area:  

They will first look for obvious physical signs of damage, then try
booting it.  If it doesn't "light up" it gets a new battery and another
try.  If it's a screen fault, it goes on the "Beyond Economical Repair"
pile.  If it's a hard drive fault, they change out the drive and try
again.  If it still doesn't work, it goes on the "BER" pile.  

At each stage, they're looking for reasons to reject the Warranty, and
the monkeys in the "Returns" department would definitely class a change
of OS as "user damage"!  If they can't get it working, and it's not
obviously due to the user, the whole thing is replaced.

> What they appear to have done is to ask the end-user to do this triage 
> test, I suspect that re-imaging the disk clears up a significant 
> proportion of all faults. Most users can't tell which faults are in 
> hardware and which are in software. HP can't diagnose the user either, 
> to tell whether they really do know what they are talking about.
> 
> I'm in two minds about whether it's reasonable to ask the end-user to do 
> this testing or whether it's better to get the unit returned and do it 
> as the first step of triage in the repair centre.

In >90% of cases, they wouldn't trust the customer to even answer a
simple "under what circumstances did the fault occur" correctly, much
less change a hard drive, so their most efficient way of dealing with
customers is just to request return of the "faulty" unit. 

> If you decide to do it in the workshop you get some additional problems. 
> Firstly a significant number (I'm assuming) of machines get returned 
> when the users could have fixed them themselves. More users have to 
> spend more time without a machine. Every user has a slower turnaround 
> for repairs. More machines getting shipped around means that the average 
> cost of a repair goes up.

The "cost of repair" is in direct proportion to the failure rate and the
original wholesale price of the equipment.  The "cost" of Warranty
Returns is factored in to the retail price, so that they (should) cost
the company nothing.

> Overall I think it's best for HP to take the honest approach and simply 
> tell all of the users that they must re-image and re-test before 
> returning any machine because that's the only way that HP can remotely 
> diagnose a hardware issue. For all I know that's what they did in this 
> case, but reporting that would have ruined the effect of the article.

That may be possible in the case of a Linux user who's obviously savvy
enough to actually install an operating system.  However, the
slack-jawed Windows "using" majority probably have enough trouble just
finding the power switch!

The first principle of "Sales" is that "The Customer is always right".
The first principle of "Returns" is that "The Customer is invariably
wrong and usually stupid".

Chris


-------------- next part --------------
-- 
Gllug mailing list  -  Gllug at gllug.org.uk
http://lists.gllug.org.uk/mailman/listinfo/gllug


More information about the GLLUG mailing list