Recently, I've been having a lively exchange of correspondence with my friend Charlie Hall, a manager at G2News.com, which publishes ClieNT Server News, LinuxGram and Penguin Hatchery, and OnLine Reporter, all weekly publications. Charlie has been pointing out how difficult it is to upgrade a PC and how this difficulty is, in itself, a substantial barrier to the upgrade process. He can say this much better himself.

"Do you know how much time we lost doing all of those upgrades? I remember one 6 hour session with a Microsoft tech supporter over problems with going from Win 3.11 to Win 95. And another session with a Compaq tech supporter who was terminated during the 3 hour phone call.

So, why upgrade now? [at Windows XP] We have something that works. And we know that Microsoft is capable and willing to release software that doesn't work reliably." [I would guess Charlie is referring to the industry legend that Microsoft's software is often best at and after it reaches Release 3.0.] 

We have to admit Charlie has hit on some very important points here. We asked his permission to incorporate some of his comments in this week's newsletter. Think of this as an Open Letter to PC Hardware and Software Vendors. A combination of Users Laments (Charlie's, mine, and dozens of others we've encountered), Buyers' Perceptions, and the Realities of Selling Technology in 2001.

An Open Letter to PC Hardware and Software Vendors

-- Your machines are much harder to use than you think they are. Vendors keep telling us how much easier it is to use a PC now. Usually I hear that pitch the day after I've lost four hours trying to figure out where a file went -- or how to fix an application that up until yesterday worked perfectly fine. It's always a vital application or an important client file.

-- Developers and designers (hardware and software, both) are peculiarly unsuited to understanding what mere mortals find usable. They, after all, like to program and see the world from a very different point of view. For instance, Dave Winer, an important and influential developer, feels strongly that it's very important to developers and users that Microsoft not be permitted to bundle its browser with Windows, to prevent undue influence and encourage competition and diversity. He told me

"There is some pain for users in this plan, but no pain no gain.

I think they'll start seeing benefits almost immediately. A Web freed from protecting Office will be a lot more powerful than the largely one-way-Web we have now. 

It's not that more competition is bad -- it's not, but rather that a highly skilled developer has no way of understanding (a) how much pain the users would feel; or (b) how much pain they would be willing to put up with and (c) they may not care nearly as much as he does about the issue of competition (although if we could explain it in terms of the possibility of future fewer choices or higher prices, as in airline flights, they might be able to)."

-- Users like predictability and stability. When we get our machines to a reasonably stable state (even one where we know that opening three Internet Explorers simultaneously will always crash the PC), we like it. We know how the machine behaves, we understand its quirks, and we can get along. Up-grading not only costs the money to buy new software and/or hardware (often both), it also requires us to hold our breath, sacrifice a goat to the gods, and keep our fingers crossed, all at once. We're never sure what is going to happen. 

(1) We're never sure exactly what to do. These are complex systems and we are usually running lots of software. Who can be certain exactly what to do or in what order to do it in?

(2) We've made big investments (generally in our own time and efforts, but sometimes in dollars spent with consultants or internal resources), getting our machines just the way we want them. Every preference is set. All our contacts are in the address book. All our files are in the right folders. Upgrading can upset some or all of this. At the very least, it will require us to thoughtfully consider how we are going to recreate our personal environment all over again.

(3) We're not sure what benefits we'll get from upgrading. Most of us use our personal computers mainly to receive and send email and to surf the web. Perhaps we run a personal productivity application or two. If all these work well on our current system it's hard to understand just what we'll get by upgrading. Advantages are often stated in terms that are designed to appeal to techies (or game-playing 13-year olds), but difficult for the rest of us to decipher.

-- IT departments like predictability and stability even more. That's because they must multiply the problem of an individual user by 100, 1,000, 10,000 or more. This means they like to move large groups of users from one platform or software version to another at the same time (although this really means migrating them in some scheduled way -- you can't change what 10,000 people do in one day), to minimize how many different versions of the same thing you must support. There's a constant tension between IT and users in even the best run and best liked IT shops because IT would like to standardize things as much as possible and users, bless them, are rugged individualists who are forever wanting to add one more thing to the mix. Sometimes it isn't what the users want to do (that's mainly a problem with skilled and knowledgeable users) but rather with what the users don't do or don't want to do (which is more of a problem with the mass of users who like most automobile drivers don't want to know how to fix the car or how the engine works -- they just want to drive somewhere. 

For all these reasons, IT shops try to minimize and manage change. This means:

(1) They try to make changes, especially major changes less frequently rather than more frequently.

(2) They try to schedule changes to make them as non-disruptive as possible. This means keeping them away from busy time periods (monthly, quarterly and year-end closings, for example), and scheduling them into "slow" periods (harder to find these days, but traditionally summer, holiday periods, middle of cycles).

(3) They will choose to skip potential upgrades where the benefits look small compared to the cost, including the cost of resources and disruption (not just the cost of buying hardware and/or software). They will often skip upgrades where there is little or no user pressure to provide or support the upgrade. 

We can see the results of all of these personal and corporate perceptions and priorities in the marketplace.

-- We are seeing far fewer opportunities to sell new PC hardware each time a new performance upgrade becomes available. The value of the incremental upgrades is simply not that interesting relative to the applications most users want to run. This leads PC manufacturers on a silly and futile search for "The Next Killer App," which to us looks a lot like the search for the lost chord. Killer Apps are never built to order. They just happen. We don't know they're Killer Apps until after millions of users have stood up and shouted "I can't live without that!" And they can't do that until they see the new something that doesn't exist yet, so there's no point in asking them what they'd like because if they can describe it the application probably offers only incremental improvement or is only interesting in a niche market -- and that's NOT a Killer App.

-- We could change this by making PC upgrades much easier to manage. That would have the effect of making the cost of migration much lower and changing the value proposition for the user or the corporation. But we can't do that as long as PC makers think of easy as an incremental proposition, judged from their own, largely technocratic, point of view. One of the reasons that there's been a niche in the marketplace for Apple computers (and that their users, in spite of how hard Apple makes it to do business with them through flip-flopping distribution and support policies, higher prices for systems and peripherals, and scarce software continued to purchase another Apple when they're ready for a new computer) is all about not just ease of use but EASE OF MIGRATION. Of course, Apple is about to kill that golden goose, too. We've just migrated from 17 years of MacIntoshes to a Mac running OS X (Unix) and while we did manage to get it up and running (more about this in another article after MacWorld), nothing works the same. 

This is what we want. If PC Vendors -- Hardware and Software want to shorten their time in the lower/flat sales Hell they find themselves in (and avoid it in the future), this is what we think they need to do:

1. Assume that the person performing an upgrade doesn't know ANYTHING about computers other than how to turn one and use a few applications. You'll need to tell them everything else and they're not very patient.

2. Make it dead simple. Essentially, I want to look at one or two screens written in plain English, make some choices, press a key or mouse, and have the computer do the work. I don't want to play CD roulette. I don't want to be asked questions I don't know the answers to. I don't want to be frightened. And I most certainly do not want to be put into a situation where my data's security is threatened because I don't understand what you're telling me to do. 

3. Give me choices. Not all of us do things the same way. Some of us like the idea of a clean start on a new machine. We just want to move our data over and start using the new software the computer store, direct source, or corporate IT loaded. Actually, we'd like them to move our data over for us, please. Others want to move all their preferences and settings without taking hours to do it and then not getting what they expected anyway. Yes, we know (some of us) that there are software packages and hardware connectors that help do some of this but none of them are entirely satisfactory and they are fairly complicated to use. Remember, it needs to be dead simple. Besides, if I just paid you $1,000 or more for a new computer, I don't expect to pay another $50 for] a special software package to transfer my stuff.

4. If you're a software vendor, please remember that I WILL NOT UPGRADE JUST BECAUSE YOUR REVENUE MODEL WOULD LIKE ME TO. I need a business reason that speaks to my needs, not yours. That means substantial new functions or substantial new reasons that I can readily relate to, expressed in terms that I can understand. "More stable than the current release" is not a good reason. It implies there's something wrong with the software I bought from you last time and that you failed to tell me that so now I'm not sure I want to trust you. Furthermore, what does it mean? Will my computer require 1 hour less per day for administrative care and feeding? One hour less per week? If I can't figure it out, I'll probably pass.

And while you're all thinking about the brave new world of Software as a Service think about how to better support the upgrade process via Services. I'd be perfectly happy to participate in a service whereby my data (which I back-up with a web-based storage service every day anyway) could be downloaded to my new machine directly from that site. Perhaps you could figure out how to do that for software and preferences, too, while you're at it. 

For corporate users, plan to provide them with better reasons for upgrading. Here there are resources to do the work (or money to buy outside resources), but whether to spend it will be a business decision and you'll be competing with other business investments. You'll need to have a better reason than "it's time to upgrade," or we'll be continuing to use the computers and software we have for longer than you'd like us to.

Home/ Search / 2005 Articles / Issue Archive / Free Newsletter

Entire contents © 2001  by Amy D. Wohl. All rights reserved. Reproduction of this publication in any form without prior written permission is forbidden.