Yesterday's post talked about the great progress that Microsoft is making towards a company-wide focus on capital-D Design. As a brief addendum, I stumbled across the Microsoft Design Center site while preparing today's post and thought it was a great compliment. I particularly liked this quote, from the Culture page:

Good user experience is now common in the consumer space, and it's the next domain of differentiation in the enterprise.

We saw a lot of evidence of their new priority during the two days and there was a lot to be proud of. Today's post takes the opposite approach, focusing on the places where progress still needs to be made. This report is centered around the statement:

Design doesn't matter enough (yet).

In all fairness to our hosts, they're at the beginning of a very long journey to get a 30,000 person company to change the way they've always done things, in a culture not traditionally known for its attention to the needs of users. It was a fantastic opportunity to be involved and to be asked for my input and I hope that today's post helps to focus future efforts.

Some of you may have seen this corrollory to yesterday's post coming. For all of the discussion about the importance of UX and design, the Round Table was also rife with examples of things changing and yet remaining the same. You have to remember that Microsoft is a big ship and big ships turn slowly. The term "lipstick on a pig" got thrown around a lot during the two days, sometimes in reference to Microsoft not wanting to keep applying cosmetics on porcine products, sometimes in reference to the continued tradition. There were two incidents which particularly stand out in my mind.

Gadgets: Do as I Say, Not as I Do

During his talk about Gadgets in Windows Vista, Michael Suesserman showed us a Gadget he had built which interfaces with Virtual Earth. He had mentioned earlier that there is a certification process for Windows Live Gadgets (though I couldn't find any info online), so I asked if that included checking their adherence to the official Microsoft: User Experience Guidelnes for Gadgets, which would indicate a firm committment to UX at an organizational level. The answer was, unsurprisingly, no; certifcation is at a code level to  make sure that the Gadget does nothing malicious. The Guidelines are pretty explicit in terms of what controls should look like and how Gadgets should be laid out and it was disappointing to see that the Virtual Earth Gadget failed to follow most of them. I pushed Michael on why his widget doesn't and he replied that it was just a quick thing he had whipped up as a demo. I think that statement summarizes some of the challenges that Darren and crew will face in their quest: many developers have a natural tendancy to view UX as a waste of time or as unecessary (lipstick on an already very attractive sow?). If they have no problem using their products, why should anyone else? I suppose that approach is okay if you're building a little demo for your own purposes, but Michael's Virtual Earth demo is the official Sidebar Gadget Tutorial (.zip file) available for download from Microsoft's Gadget Developer site.

Why is this Bad?

This sends a message, from up on high, that no one else needs to bother with the Guidelines either (in all fairness, Apple does the same thing by continually demonstrating their determined refusal to follow their own User Experience Guides by releasing applications which violate them entirely).

How can Microsoft Fix This?

It's a pretty easy one to fix: give the Visual Earth Gadget to a designer to clean up and add some content into the Tutorial which explains why the Guidelines are important to follow.

Know Thy User

I haven't talked about Expression yet - that's tomorrow's post - but there was some confusion around Expression Blend which is relevant to today's topic. Without stealing my own thunder for Thursday, Blend is intended to end the interminable battle between Designers, who wear shirts that say "#000000 is the new black", and Developers who wear shirts that say "foo() happens;" - a rather clever way to say that it helps everyone to speak a consistent language. The goal of the app is to allow designers to build controls using drag-and-drop and simple scripting instead of handing developers specs which they won't follow. Once the controls are built, the developers can pull them into Visual Studio, integrate with the backend, and be done with it. Alternatively, in what I thought was a much more powerful (and real world) demo, Arturo went the other way and pulled a really ugly app from Visual Studio and reskinned it without breaking the functionality.

After the demo, the question was thrown out to the room about how we would use the app in our own practices. There was some confusion about what Blend actually does, which was amplified by what the differences are between WPF and WPF/E, the codename for the Everywhere version which will run on Macs, handhelds, etc. The longterm outlook for WPF/E (thankfully to be renamed very soon) is that it's a competitor to Flash (and specifically to Flex and Apollo). At the moment, it lacks any sort of interactivity and is limited to playing back media, so it's really not particularly useful. The answer to the question, at least from Radiant Core's perspective, is that we wouldn't use Blend because we're not currently in the business of building Windows applications and we couldn't reasonably deploy client projects in WPF/E in its current state. What was more interesting were the answers around the rest of the table: there was much confusion about what kind of "Designer" Microsoft had in mind to use the app. We all agreed that Visual Designers would rebel en masse if they had to start actually building controls, and that their decisions about how to build them would make Developers rip them up and start from scratch. I suggested that it might be useful for the people in the room - UX Designers - to be able to wireframe a project in Blend at a very early stage, then share those wireframes with Designers (who could use Expression Design to make the visuals, export in XAML, and then import to Blend), and Developers (who could start planning their implentation by importing the Blend XAML into Visual Studio). There seemed to be agreement around the table that people could use it that way, which seemed to be news to the Microsoft staff in attendance. In the end, Blend seems to (in a broad sense) be a replacement for VisualBasic, so the whole marketing angle of empowering Designers to build applications seems somewhat misguided. The whole conversation made me wonder if this was the first time that Microsoft had really asked people what they thought - in other words, if this was really the first UX activity which had been conducted on the product.

Why is this Bad?

The very first rule you need to know about successful UX: know thy user. Aiming a new product at the very broad market of 'designers' reminds me of Ali G's pitch to Donald Trump (and various VCs) for ice cream gloves (note: not everyone finds Ali G funny or safe for work so play carefully):

Sure, almost everyone likes ice cream and has hands, but that's not really a demographic you can market to. What do they mean by designers? From the Wikipedia article on the term , here's a selection of people who might be included:

• Automotive Designer
• Costume Designer
• Fashion Designer
• Jewelry Designer
• Game Designer
• Graphic Designer
• Industrial Designer
• Interior Designer
• Landscape Designer
• Scenic Designer
• Systems Designer
• Web Designer

I'm belabouring the point a little (though I did get to sneak Ali G in), but hopefully you get it. 'Designers' in no more a market segment than 'Canadians' or 'Dogs' or 'People who drive cars'.

How can Microsoft Fix This?

This one is much tougher, since Blend is basically ready to ship. In an ideal world, they would have involved their potential users from the very beginning to make sure that they were building something interesting to them (and they may have done so and not made that clear during our session). To fix it now, they should figure out who would find it interesting (UX folk are definitely part of that group), and then target their marketing and sales efforts at those people. Rather than the loose 'designers', focus on specific disciplines within design for whom Blend solves a real pain point.

That's it for Day 3 - tune in tomorrow for an overview of Expression!