Back in March of this year, Frank Hayes a columnist for ComputerWorld (one of my favorite weekly magazines), wrote a fantastic article called “Innovate Big.” I loved this article so much I cut it out of the magazine and posted it on my office door for everyone to see. We can related to this article in many ways on Windows. We look at this challenge as a balancing act. There are many legacy components in Windows were we want to update the text, but many times we shouldn’t touch that because it causes frustration for our users. Yes, I’ll be the first to admit that we’ve definitely went ahead and made minor changes which goes against these principles, but generally we try to stay away from it.
The challenge for Fit & Finish is that we want to polish UI over time but can we do that without affecting user’s habits in a bad way? Also, does the change out weigh the cost (for making the change & the user’s learning cost)?
Many people don’t realize this but these changes come with a surprisingly high “tax” that goes into updating Icons, changing text, replacing a graphic, etc. I’d like to share with you what goes into some of these changes:
“Taxes” for making a small change:
- Someone needs to prepare the change
- Text change – The writer needs to write the change and an editor needs to approve it.
- Icons – The Icon designers need to make the Icon and provide all the right Icon sizes.
- Graphics – The graphic designer needs to design & create the graphics.
- Someone needs to make the change
- If a developer checks this in, it takes time away from other issues.
- Test needs to verify the change
- Verify the change is correct before it gets “checked in.”
- Verify the change is correct when it shows up in the build.
- Localize it (the most expensive “tax”)
- Does this change require any localization?
- Do we need to provide any additional resources for each locale?
- Accessibility
- Does it affect Accessibility in any way?
- Does this change break the High Contrast experience?
- Does this change break the High DPI experience?
- Geopolitical
- Are there any geopolitical issues with this change?
- Do we need to provide other resources for other regions so it’s geopolitically safe?
- Legal
- Are there any legal issues with this change?
As you can see, there really is more to it than just making a quick update. You’ll notice that we did “pay the tax” in many areas, especially with Icons. There are other pre-existing conditions we need to think about especially for legacy (already shipped) components:
- Will we interrupt the user’s workflow?
- Will the new change cause additional confusion?
- Does this affect any existing help documentation for Microsoft or any third party documentation?
- Is a Knowledge Base (KB) article required after this change?
- Is this a change worth making?
Quoting directly from Frank Hayes’ article, Innovate Big about small innovation:
“Users hate that. And they're right to hate it. Broken habits aren't just an inconvenience. Habits are part of a user's work process. And a smoothly running process means productivity. Break users' habits, and you've broken the process, spiked productivity and forced users to develop new habits. That takes weeks at best, and months if the old habits are deeply embedded.
That's the price of small-scale innovation. It's a price that can be measured in lower productivity and higher training and support costs, user dissatisfaction and damage to IT's reputation. Look, IT people like little innovations. They're easy and cheap -- for us. But for users, those small, cosmetic "improvements" can be pure misery. And too often we ignore just how much they really cost.
Is there a benefit to your innovation so big that it's worth that price? Then go for it. Never underestimate how disruptive small changes can be, but never let that prevent changes that are really necessary.
But if there's no real benefit to a small change, don't do it. In fact, work hard to keep those small-scale, user-level details the same. You'll have happier, more productive users and an easier time supporting them.”
Let’s look at an example of a small text change. The “Computer Name” tab under the “System Properties” dialog has always bugged me. I’m not a fan of the text that we display to “Join a domain.” We could easily change the supporting text and the push button text to make more sense, but what value do we add if we change it?
It’s still the old cruddy interface; would we really improve anything if we change the text? In my opinion, no, we wouldn’t. There are many people that are already “trained” to use this UI and it would just be a disruption for them. Sure, there is probably an inherent affordance for users that do this frequently, but how often are you joining a domain (although if you work on Windows you do this everyday 
)? The pain here is for product support folks out there that need to go update screen shots, and help documentation referring to how to join a domain. Yes, you will need to do that anyway for many of the features in Windows, but why add the hassle for this legacy UI component?
Vinny Pasceri
Aero Program Manager