Fog Creek Software
Discussion Board




IT Users: Colateral damage? Who cares

It is interesting that Joel brings up the comparison with the telco deregulation: "nobody wanted to use Sprint or MCI, because their service was terrible, but we were sure thankful that reliable old AT&T had to lower their prices in response."
I am pretty sure that "we" did not include the users that were forced to deal with that "terrible service" on a daily basis at that time.

Things have not changed, as a bit of browsing the "Munich report" will show.

This report tells us in numbers just how much they care about the user. As we will show: not a lot.
At the end of the report there is a sort of cost/benefit analysis. They divide projected cost by a perceived "strategic-quality related intangibles". This "quality" is not the technical stuff you might expect, that is dealt with elsewhere. So what does it contain?
Well, for one thing it is here that the effects on the end users are calculated, both the people working in the services of the municipality as well as people interacting with the municipality.
Let's look at the numbers: The MS solution literally obliterates the OSS in the case of the employees, getting a score that is almost 7 times higher! Now that must surely be devastating right? Well, not really.

What they did is clever and a bit subtle, but hey, we are all intelligent subjects here right? So bear with me.
Officially the "influences on the employees" figure counts for 19% of the total score. Ok, slightly less than 1/5th, poor but bearable. Now surely the scores given to these things are a bit of handwaving. Surely there can be relative interpretations like "solution A is really very far ahead from solution B in this criterion". Making the stuff stick absolutely becomes more questionable. How "perfect" is solution A?
What they did was give all the solutions a relatively low score on user issues: in fact, the best solution (MS in this case) only scores 43%, where the OSS scores a very meager 6.5%.
(For the external users the overall criterion weight is 5%, with maximum scores of 26% (MS) vs. 13.2% (OSS).
Now cleverly they do not give out low marks across the board. E.g. on the "security" criterion, that already carries the highest weight of 29%, the give out scores of 93% (OSS) and 83% (MS). (Before you MS people start debating, keep in mind that the municipality itself admits their security practices could be improved for MS platforms. E.g. they do not use any normal MS security related server products for management (no AD, no SUS, no Group policies, nothing)).
Now this is strange? Let's see. What is the picture that emerges when we "normalize" the criteria to the highest mark given?

Before normalization we see:
- Upholding the rules: 14%
- Security:29%
- Employee satisfaction: 19%
- system management: 9%
- External users: 5%
- Politics and Strategy: 24%

After normalizing the scores we see the real resulting influence:

- Upholding the rules: 11.4% (- 2.6%)
- Security: 38.4% (+ 9.4%)
- Employee satisfaction: 11.7% (-7.3%)
- system management: 8.6% (-0.4%)
- External users: 1.9% (-3.1%)
- Politics and Strategy: 28% (+4%)

So instead of the given 24% influence of the end users, there is really just a meager 13.7% taken into account.

(yes, I know the scores for most of these things are bogus, especially when they compare a badly managed system with a potentially perfectly managed hypothetical system etc. but that is not my point here)

Conclusion: Yes, in 2003, we still do not care about the users. Usability? Who cares.

Just me (Sir to you)
Thursday, July 17, 2003

I fail to see how usability can be quantifed before the apps have been ported. If you're just doing word processing or receiving emails then you will scarcely know what OS you are in.

I think it's a political decision, and a correct one, that even Windows users should be happy about.

Stephen Jones
Friday, July 18, 2003

I can see this is going to be a loosing battle, and of course no one to blame but myself since I tried to make the point given a highly sensitive example.

Forget the whole Linux vs. Windows stuff.
No. Try again please. Forget about it just a second. This is not what this is about. There is a whole topic devoted to that stuff here http://discuss.fogcreek.com/joelonsoftware/default.asp?cmd=show&ixPost=57265 .

Ok, now say you are imagining the aquisitioin of a cookie cutter, and there are two altenative offers. Among the many things that will influence your decision there is the matter of the daily experience of your thousands of employees operating the cookie cutters. You can only make a guestimate, but you guess that machine A will give those people an "I'd rather poke needles in my eyes while someone is smashing blocks of concrete on my little toes", and machine B will give them an "almost bearable but nothing to be thrilled about" score.
The question now is how much this is factor going to influence your descision?
Some evidence suggests that at least for IT aquisitions the answer seems to be "very little".

Does this mean that an investiment in improving the usability of your products when you are targetting a market where the purchase descision is not in the hands of the final end user is just a waste of resources (or maybe it should be recognized as an altruistic act of pure compassion)?

Just me (Sir to you)
Friday, July 18, 2003

Take the example away and maybe you're right.  We have a database at work that is so bad that the staff are divided into two camps; those that enter the correct marks and attendence, and those that don't give a shit. Among the first there are a significant proportion who think the UI is so bad they would change job because of it.

Stephen Jones
Friday, July 18, 2003

This issue is not very clear to me at all.  It does sound like employee morale and usability have been disregarded, but I am not convinced that it's quite the problem that it sounds at first.

One working principle of IT is that humans are almost infinitely malleable and people who first encounter a computing tool that feels quite clumsy generally can't imagine a near future in which they will become accustomed to that tool. Everything in this business sucks when you first use it: mice, PDAs, Windows itself, the Control Panel, etc. I suspect that this 'understanding' was implicit in decision making here.

We all know as developers that this happens again and again. People bitch about change, but adapt, and generally prosper in time. Employees don't all of a sudden go homeless and hit the streets, or go into deep clinical depression, because their programs launch in a different way. It happens whenever a developer first switches platforms - one year later they won't go back to the old platform. And computer users have coped with ugly green screens for years simply because there weren't any alternatives.

While a Linux KDE type desktop can be and is a nerve wracking, grating transition to make from Windows, I would bet that most normal computer users will be able to learn enough to do their jobs. I think that for long term employees using these desktops, the initial transition will be highly irritating, but longer term, they will not notice the differences so much.

The *real* productivity and morale hit will come with people who need to use these systems occasionally, such as temp workers and visiting professionals and other employees. Another hit will be those who switch between Windows at home and Linux on the job.

One thing clearly has to happen for Linux to gain ground on the desktop: Linux developers have to pay *excruciating* attention to every user interface behavior and decision  made in Windows and offer a suitable emulation in Linux. Only then will transition to Linux not  be characterized so much as a hardship.

And common hardware settings like desktops, mice, multimedia, etc have to stop being an afterthought. IE: what is out there is an excellent start, but for instance, how the hell do I set up a desktop shortcut in Linux? Or how do I *easily* (not being an Xfree86 expert) change screen size? Does sound support work and can an end user configure it? Etc. Few in the developer community seem to regard "feel alikeness" as a worthwhile goal, since it's not a cool internal.

Bored Bystander
Friday, July 18, 2003

*  Recent Topics

*  Fog Creek Home