30 Comments

Summary:

Microsoft says it has seen the light in terms of designing software that business users actually want to use. For its new Q&A feature for visualizing Excel data, for example, the product team spent six weeks thinking about UI before even thinking about technology.

olympic medals

A few weeks ago, I highlighted a new data-analysis tool from Microsoft that automatically analyzes and visualizes data as users type into a search bar. The feature, called Q&A, is an impressive piece of technology, hiding some complex computations under a deceptively simple user interface. That was no mistake.

“Microsoft was not a design-first company for many years,” explained Microsoft Technical Fellow and Q&A team member Amir Netz during a recent demo of the product. “… You see the design-first [mentality] now permeating even the highest-end enterprise products.” (We’ll be featuring the best in experience design at our RoadMap conference next month).

The past: Let’s call it suboptimal design

What he means is that Microsoft was once — pretty clearly — guilty of the classic feature-first business software mindset. Throw in everything a power user might want and largely ignore the fact that most users don’t want to wade through 90 menu options. They don’t want a toolbar that bears a striking resemblance to the TV screen when Fox Business is on. In an office of 20 people, it doesn’t matter if a product can support 10,000 users simultaneously.

Uh, what am I looking at?

Uh, what am I looking at?

Speaking specifically about business intelligence, or BI, software like Q&A, Netz said the status quo is that users normally need strong skills in manipulating and analyzing data, and “a willingness to get dirty.” Because of the barrier to entry (as well as some technological and cultural issues that still persist), Microsoft GM of Product Management for SQL Server Eron Kelly added, it can take weeks to get an answer to a simple business question if it means someone has to go through and reconfigure the report.

Customers waiting on new versions of software could expect to wait 3 years in between releases. “If you were really aggressive you could get it to 2 years,” Netz said.

Add in the process of deploying traditional desktop software, which might mean buying and installing numerous different packages just to get started, and you get what Kelly diplomatically called “friction for users to consume our technology.”

The future: Simple, fast and addictive

With Q&A, however, Microsoft “started with just a pure design view,” Netz said. That meant the design team sat in a room for weeks and thought about how to make the user experience as simple and addictive as possible. They worked around the clock and on weekends, he said, with the understanding they’d move on only when “everybody in the room feels that what we have is just going to be awesome.”

That happened about six weeks later.

Only then did the team begin thinking about the technology and the architecture that would underpin the aesthetic and interactive experience they’d imagined. They brought in other Microsoft personnel with expertise in search, BI and natural-language processing, who were tasked with making Q&A intuitive, intelligent and responsive (it’s powered by an in-memory column store that Netz called “maybe the fastest in the industry.”). A prototype was ready about two months later and “was way better than what we expected,” Netz said.

Fast-forward another 14 months, to September 2013, and the full-time that Microsoft assembled to build Q&A had the product ready for its public beta release. Like most cloud services, it will receive updates monthly. And whereas the old-school BI software really targeted about 20 percent of business users, Netz said that with Q&A, “We think there’s a high potential to reach the other 80 percent we’re not even aiming at.”

But did Microsoft deliver?

There’s still work to be done on Q&A, but it’s difficult to find much fault in it — at least with the very curated demonstrations that I saw. Netz and Kelly easily sorted through a spreadsheet full of Billboard music data, showing various ways of sorting it by songs, artists, decades, genres and any other category included in the data.

bestartistqa

By “best artist,” which the software decides means weeks at No. 1.

qagenre

The number of songs that charted in each year, by genre.

I experimented with a preview version that included preloaded datasets about the Summer Olympics and sales data from a hypothetical local bar. It was easy enough to search for the recipe for a particular drink:

sexrecipe

And figure out on which days it sells the most:

sexQ&A

Or to sort a map of Olympic host cities by the United States medal count in each (Q&A automatically switched to the Olympic data from the bar data based on my query):

olympic medalsAnd then to switch to a column chart from the map and to open up a view into the underlying data in case I wanted to do something more advanced:

olympicmedalscolumn

After all, Netz explained, Q&A is essentially just an easier way to search and visualize data that’s already stored in Excel.

If there’s nit I have, it’s on the business side: Microsoft’s maze of services and add-ons still isn’t the easiest to navigate — Q&A is a component of PowerBI, which is an add-on to Excel, for example, whereas Tableau is Tableau — and tools like PowerBI are clearly targeting business users. It might democratize data analysis within companies, but not among everyday consumers who want to quantify their lives.

But if Q&A really is indicative of where Microsoft’s software design strategy is headed, it would seem, by some measures, that business’s best days could actually be ahead of it. Office 365 might not dominate the cloud like Office dominated the desktop, but the users who find themselves in front of Microsoft’s services might find themselves a lot happier.

  1. They need to reconsider their pricing for SOHO. They’ve seemingly abandoned that market.

    Share
  2. BenjaminGilead Monday, October 14, 2013

    That’s a problem with a company with a salesman as a CEO. They are always following other’s company lead.

    Share
  3. johm mcdonough Monday, October 14, 2013

    It sounds like another “check your brains at the door” company, learn to trust your employees ideas.

    Share
  4. 6 weeks on design? To fix 25 years of bad interface, most recently the ribbons? Sounds inadequate and token.

    Share
    1. I agree. They should have been sent to Tibet for four years, smoking hallucinogenic plants and fasting as to cleanse the MS out of them before being asked to design anything. Including toilet paper.

      Share
  5. H. Elwood Gilliland III Monday, October 14, 2013

    Wow 6 weeks.. they’ve had 30 years

    Share
  6. Design first, what an interesting idea. Apple has been doing it since the 70’s. Bwaahahahah

    Share
  7. The bastard are ripping everyone off with Metro

    Share
    1. Really? They were the first big player to switch to flat design.

      Share
      1. 8 is not great Tuesday, October 15, 2013

        And that’s all it is. FLAT

        Share
      2. Flat design is crap, Visual Studio is almost unusable now with all the stupid changes. They should have listened to the developer community before the tried to force that abomination on us. Complete, utter garbage.

        Share
  8. Just Some Guy Monday, October 14, 2013

    So they’re going to try ready, aim, fire? Wau.

    Share
  9. Crystal Noe, M.S. Monday, October 14, 2013

    Sounds very interesting and I would love to try it. In most offices, there aren’t BI specialists, but everyone needs tables, charts and heat maps. SAS isn’t user friendly or affordable for most non-techies. Tableau, although ultra user friendly UI and free training videos, requires your data be BI ready for the most part. This, can be difficult for non-data folks. Qlickview is versatile with data prep, but not user friendly either. Excel’s BI ribbon requires the user to have SQL installed, but most companies won’t approve SQL for a non-tech user. Further PowerPivot wasn’t user friendly and its web site was down half the time. I would love to see an easy tool for everyone in the office. I look forward to checking it out. (For the record, why are people discussing Apple?? This is a BI app, Apple isn’t in that space and hasn’t created a BI tool for their users…..the comparison makes no sense? 6 weeks? Sounds like an agile approach with focus on market needs and time to market. Seems normal to me.)

    Share
    1. What a sound, logical, common sense average business user response – – wonderful! You hit the nail on the head on every aspect. If MS can get a BI product into the hands of everyday workers by building a great intuitive, value add product, I say BRAVO!

      Share
  10. All the pathetic Apple fanboyism in these comments…you sad people need to get a life.

    Share
    1. Samuel A. Falvo II Monday, October 14, 2013

      Dick, there’s 11 comments as I post this. There’s exactly 0 comments that are in any way Apple fan-boi-ism. Zoltan’s comes the closest, but as it’s factual (and can be backed with evidence), I don’t consider it a typical fan-boi response. Which begs the question, do you even know what a fan-boi is, or are you merely preparing for some future blast of comments yet to happen?

      Share
      1. If you insist on using the phrase “begs the question” please use it appropriately.

        Share

Comments have been disabled for this post