The fact is, many features are a liability. They add more things that can go wrong. Of course, if your need for a particular feature outweighs the liability it carries, then it makes sense to pursue that feature. However, often times you are forced to buy more than you need in the process. Again, motherboards are a great illustration. The charts below show the failure rates for various motherboards we’ve sold in the past, with varying amounts of features onboard (the “Deluxe” and “Premium” boards have more features). The pattern is clear.
| Intel P35 Motherboards |
Failure Rate |
| Asus P5K3 Deluxe Wireless Edition |
18.60% |
| Asus P5K Deluxe Wireless Edition |
11.43% |
| Asus P5K |
11.21% |
| Asus P5K-VM |
5.33% |
| Asus P5K EPU |
4.55% |
| Intel P45 Motherboards |
Failure Rate |
| Asus P5Q3 Deluxe/WiFi-AP |
23.81% |
| Asus P5Q-E |
8.24% |
| Asus P5Q-EM |
6.18% |
|
| NVIDIA NFORCE for Intel CPUs |
Failure Rate |
| Asus Maximus Formula Special Edition |
50.00% |
| Asus Striker Extreme |
40.70% |
| Asus P5N32-SLI Premium Wireless Edition |
41.38% |
| Asus P5N32-E SLI |
26.02% |
| Asus P5N-E SLI |
13.33% |
| NVIDIA NFORCE for AMD CPUs |
Failure Rate |
| Asus A8N32-SLI Deluxe |
20.16% |
| Asus A8N-SLI Premium |
12.50% |
| Asus A8N-SLI |
7.46% |
|
You may be thinking that these failure rates are shockingly high. The key is our definition of “failure.” If ANYTHING is out of the ordinary with a product, we call it a failure. It may be a simple cosmetic scratch. Or a chip that shows hotter than it should on our thermal imaging. Or it benchmarks more than 5% out of spec with what we’ve seen in the past. Or maybe one SATA port doesn’t work. Many of these things you would probably just choose to live with if you were building your own computer. We send it back to the manufacturer until it is right (and they grumble plenty). A vast majority of failures are caught in our factory — that’s our job!
There was a time when the call for more features enticed even us at Puget Systems. Under strong demand, we once sold the Asus Striker and Maximus series of motherboards — models known well in the enthusiast community for their exhaustive list of features. The results speak for themselves in the chart above. We quickly realized our error and corrected our product line. From that point on, we embraced simplicity. We understand that one of the big advantages of being a custom computer builder is about what we DON’T sell you. By simplifying our products, we can not only save you money, but we can create a dramatically more reliable product. Are there things you need? By all means, meet those needs! By going custom, you are better equipped to do that without paying for and exposing yourself to the risk of features you don’t need.
Of course, product reliability is only one factor to consider. Difficulty of maintenance, longer boot times, and use of system resources are some other “hidden costs” of features. I appeal to product reliability only because that is one of our core focuses here at Puget Systems, so the hard data is already right at my finger tips!
|
5 Responses to “Features are not always selling points”
Great article, Jon. This is a great contrast between features and benefits. Benefits or perceived benefits sell. Features are nearly meaningless unless it’s an engineer on the phone.
By Joshua McClure on Dec 30, 2009
Thx
Jon
By Robert on Dec 31, 2009
jon great info you have had some real good blogs lately. but i disagree to a point…simple isn’t always better but it sure can be limiting to the end user at times.
So now how about showing current i7 socket 1366 users that choose Asus P6T boards their problem/failure rates…. wether P6T Deluxe (original with SAS) and P6T Deluxe V2, and P6TD Deluxe and any other of the P6T line you may have changed with…. for whatever reason.
You must have the figures on those by now. some may have had more service calls than others, bios upgrades, memory touchiness, etc.
I’m betting that would be interesting to current users.
and to get back to simpleness for a moment it’s limiting and while you may not sell 80 units with a sound card that is more for the music listener than gamer i see more nad more users going that route. your orders may show the exact opposite as more users going with onboard sound….. blah!
while i commend you all for at least giving asus cards a shot even though they based it on previously developed chips it beats creative.
the sound card set is industry limited in this price range this i understand but there are others that are at least kept up to date by their respective companies.
you built me one stable machine… and yep i am still running vista ultimate 64 albeit via a recent clean install which has fluctuated via 10.00 and 9.88 on “reliability monitor” and i run all the cakewalk software and ableton updates along with all nikons software.
so other than adding some ram for my music work along with photography stuff. i can’t complain but would have used more choices had they been available.
By Mark Adelman on Jan 9, 2010
Great question — the P6T series has gone through two revisions since we started selling them a year ago. It is interesting to compare their reliability. Its a bit off topic from the study of reliability vs feature sets, but let’s give it a look! I would expect each revision to be incrementally more reliable, as Asus fixes problems found in the field.
Asus P6T Deluxe (began selling Jan 09): 28.26% failure
Asus P6T Deluxe V2 (began selling Mar 09): 5.64% failure
Asus P6TD Deluxe (began selling Oct 09): 6.82% failure
The 28% number looks alarming, but on close inspection I see that about 60% of those failures were due to a defective battery mount — the mount in that model was not very well designed, and units would arrive to us broken. We actually were one of the early reporters to bring that to the attention of Asus, and they fixed it in the V2 revision. Only 5% of those failures happened in the field.
Even so, you still see a reduction of failure rate starting with the V2. The P6TD sees a slightly larger failure rate, but it is within our margin of error, so I wouldn’t read too much into it.
I didn’t mean to imply a blanket argument against features — just ones that aren’t needed. If you need a feature, absolutely, go for it! My encouragement is that people should think about what they *really* need. Especially with our market segment. Some of our customers are quite well off financially, and I think the discussion needs to be about more than the monetary cost of features — there is a reliability cost as well.
By Jon Bach on Jan 9, 2010