Transitioning to Office Labs


It’s been about a month since we went live with officelabs.com. Of course as a team we’ve been operating a little over a year, but only now do I get to talk about that period. I thought I might fill you all in on my transition from OneNote to Office Labs.


As Office 2007 was wrapping up, my second son Skye was born. (I’m sitting next to him on a plane to Toronto as I write this).  I took my paternity leave and when I came back in the fall of 2006 Office 2007 was just wrapping up. I went to work for Jeff Raikes, the president of the Business Division at the time. He had asked me earlier to come do a job for him whenever I was able to move on from Office2007.


Jeff had a two-part mission for me that was simple to say and hard to do. Basically he said, “help the division try more ideas”, and “explain to the world and the company what our long term vision is for productivity”. Right, roll up those sleeves and start knitting Petunia!


For background, the business division covers several product lines: Office System (which includes the Office applications, SharePoint server and a lot of other components and servers), Office Live  – the productivity service, Unified Communications (UC) which includes Exchange, Live Meeting and Office Communicator (enterprise IM), and Microsoft Business Solutions (MBS), which provides “software to run businesses” such as ERP, CRM and several other three letter acronyms . Basically the business division as a whole is focused on “productivity” – helping people get stuff done  – whether it be at home, school, or work.


I worked in the Office group for 12 years, first a little on Excel, then a lot on Word, all of OneNote, and a little on Publisher. I was quite familiar with the Office team’s processes for developing software which are in my opinion first-rate. Relative to just about any software project in the world, the scale of a release of Office is huge, the quality that comes out is excellent and always improving, and it gets done more or less on time which is a semi-miracle in software development. But that doesn’t mean Office and its processes can’t improve – in fact the Office team is first to criticize themselves and work on ways to do better.


The other parts of our division have their own styles and personalities. Office Live is fairly fast moving, Unified Communications is always up for a new idea, and MBS is consolidating  and growing its scope at the same time. Jeff had asked me to help the division “try more ideas”. Ok, where to start?


One thing that was clear was the different viewpoints on “innovation”. Some people in the company felt that we didn’t explore new ideas and technology well or fast enough. Others felt we had no problem and that we took the right measured approach to new things. I think many readers who work in corporations are familiar with this range of opinions about their business. Sometimes it can get a little nasty – the “gotta innovate” people think the others are blind curmudgeons who couldn’t imagine anything other than what they have today. The “steady as she goes” folks think the others are lightweights, chasing the latest trends and not being appropriately thoughtful about business results. That’s a caricature but you get the idea.


Our popular culture definitely is biased toward the new – we tend to place it on a pedestal, and imbue words like “invention, innovator, entrepreneur” with mystical heroic qualities. People admire and many aspire to be seen as innovators. The innovator who makes a $100m/yr company is lauded. The guy who quietly makes a billion dollars a quarter doing what they did last year is ignored. Note: as much as John Q. Public admires the former, shareholders really like the second type of guy. Delorean makes a good story, but we can’t name who at Toyota made them the #1 car company in the world.


I’m conscious as I write this that the “blogosphere” tends to have a huge bias in favor of rapidly embracing the latest thing – most of us are in that “gotta innovate” group – more than even the general public. But a criticism of that approach especially for a larger company is that you can’t always be chasing “shiny objects”. You have to deliver business value. Speed for speed’s sake is also not a good idea. How do you act appropriately fast enough without just churning? After all, while consumers might say they want a new thing every day, enterprise customers tell us they can’t handle the pace of change we throw at them today. What to do?


I didn’t want to fix a phantom problem. I also felt during my time in Office that we could make some improvements to the system, as good as it was.


One thing I felt strongly was that the people who said we don’t have enough good ideas were flat wrong.  There is no shortage of great ideas at Microsoft. I also felt that our product groups, even the “old” ones, had the appetite to take big risks. Just look at the new Ribbon UI for Office 2007. We also had the capacity to develop entirely new things on our own – look at OneNote, or SharePoint. Where it seemed we could use a “tweak” was in five areas:



  1. Improving the number of risky or uncertain areas we explore at any time

  2. Increasing the “stakes” of the areas we explore – place bigger bets

  3. Decreasing the iteration time for designs (more iterations, faster in response to user feedback)

  4. Accelerate in time the delivery of new ideas (get them to market faster)

  5. Easier exploration of ideas that didn’t already have someone working in the area

We’re trying to tackle all of these with Office Labs. Our team extends the already solid Office development process to enable those product teams and others across the business division to try more stuff. When I write next I’ll go into more detail.


One closing thought. We work with the whole business division (which makes more than Office). We also aspire to create entre products and services, which may not be called “Office”.  So why did we call ourselves Office Labs? The answer is that Office is really well known and is much bigger in scope than most people realize. So it actually covers a lot of the productivity area. Trying to come up with a name that was generic like “business division labs” or “productivity labs” seemed lame. And choosing a fancy sounding name just to be cool was not our style. So Office Labs it is – Office System is our largest and most important client, and its definition is always growing – in fact we hope to be part of that growth.


Comments (7)

  1. Joe says:

    I agree with your point about chasing shiny objects.  

    But I have to say that (from the outside view) Office hasn’t changed very much in the years I’ve been using it.

    You say there should be user demand for a new feature – but the average user probably just doesn’t expect to work in a different way to they do now.

    For example, why is Excel a separate application – why can’t an Excel sheet be a table in a document (or a canvas).  In fact, why is most of office based around printed paper – when these days most documents never make it beyond the screen.  

    What about email and outlook?  When will it support asynchronous I/O operations?  How about modelling a conversation as something other than a linear email body.  How about exposing that conversation as data that can be published as a conversation –  I could go on….

  2. Chris_Pratley says:

    Joe, thanks for the comment.

    Abotu Office not seeming to have changed much. You can make a good case for that, and you can make a great counter case for it. 10 yrs ago there was no SharePoint, OneNote, InfoPath, Forms Server, Excel services. All the applications have picked up many new capabilties – you just have to switch back to Office97 and try that for awhile to notice them. OTOH, the core apps are still designed to do what they were designed to do, so certainly you could say they haven’t changed.

    I can’t find where I said there should be "user demand" for a feature in this post, but it’s a good point. I’ve written before about articulated needs (user demands), and unarticulated needs (needs people have but don’t ask for – because they don’t know they have them). The former tend to be smaller things ranging from bug fixes to feature add like say, a better way to compare two documents. The latter tend to be larger things ranging from huge new capabilties, products, or "the internet" which most of us didn’t know we needed and couldn’t have asked for before it appeared.

    There’s a saying that "if you don’t listen to your customers someone else will". There’s another saying that "if you listen to your customers too closely you’ll miss the next big thing." They’re both true of course. Listening to what customers ask for is important – you’ll lose your business sooner or later if you don’t. You also need to look beyond what they ask for or even would seem to like to be ready for the next great opportunity. Think of the horse and buggy manufacturers, listening to their customers ask for better bridles while the next guy over is building the automobile.

    Why is Excel a separate application? Is it historical, or is there a good reason? Probably both. if you ask a finanical analyst if they want to use a table in Word that had Excel fucntionality for their work they’d laugh at you. But for simple calculations that sure would be handy. In fact 20yrs ago this idea of compound documents was all the rage – that’s how Object Linking and Embedding (OLE) got started. Even today and ever since the early 90s you can paste an actual Excel speadsheet range of cells into Word or PowerPoint or wherever that supports OLE and have actual Excel there in your document. People don’t do that too often, and there are several reasons for that ranging from discoverability to bugs to usefulness.

    I do like the idea of a canvas that you can place all sorts of content on – that was part of the OLE dream too. But maybe today there are ways to make it work better. In fact we showed some ideas around that a month ago: http://www.youtube.com/watch?v=YPrfqdl55D0

    Outlook and email is a very fertile ground – I agree. Outlook supports async operations such as sending receiving and caching mail for offline use already. I wonder what your ideas are in this area?

  3. joewood says:

    Thanks for the long response Chris.

    I’ve been programming with the Office platform for a long time.  In fact I was a big user of Word 2 for windows (I think it was called) and even wrote a OLE server or two (with and without the help of ATL).  I also work in the finance industry – and yes we have those traders that would laugh at a table in Word being their risk spreadsheet.

    My point about Excel / Word is that if the platform was more modular then the engine part of Excel could run behind a canvas and the trader wouldn’t even know or care.  Our risk management spreadsheets could become the monthly risk reports without having to publish / cut’n’paste or embed data between applications.  And that risk spreadsheet would become a real distributed process.

    Your point about OLE2 (wow, haven’t typed ‘OLE2’ for a long time) – people didn’t use it I think because of performance and reliability.  It was the right idea with a bad implementation.  The problem is that it looks like the Office team suffered a case of ‘once bitten, twice shy’.  I am hoping you guys are cooking up a more approachable solution to this that involves XML, XAML and something like DataTemplates.  Maybe if Silverlight becomes a first class office citizen.

    And then Powerpoint just becomes a way of viewing the document/canvas.

    The key frustration with Outlook has to be that it is in the position to redefine email.  For a lot of people outlook and exchange is the pivot point to their working life.  At the moment everything is typless and linear.  An email chain can represent an interaction with a resolution that could and should be publishable as an artifact.  I see outlook and exchange as a system holding many different rich sets of data and interactions but only one real way of visualizing that data – a stream of messages in a grid.  I know there’s outlook forms, but again – how accessible is that technology.

    I realize that the foundation for all of this is there: –  Excel services, Sharepoint workflow and the potential for Live Mesh in a corporate environment is immense.  And that most of what I’m saying is on the plan somewhere (OneNote was a great step forward) – it just seems that (from the outside at least) the bigger the Office suite and services become the slower the development gets.

  4. Chris_Pratley says:

    Joe, your thinking is similar to what we are thinking, perhaps not in exact specifics but in spirit. That sort of "out there" or "redefining" work is what we are charting in Office Labs. There is always an element of "we tried that and it didn’t take off", but we all know that sometimes just small changes can make an idea fly, so nothing we’ve tried before is really off the books. The trick is to find how to make it take off.

  5. Craig says:

    I hope that your presentation today on TownSquare was successful. Will this project be opened up beyond just MSFT employee participation? If so, when? Is this a project that you plan to "Codeplex" (i.e. open source), or not?

  6. MadBabble says:

    Thanks for tips

    regards

    Sven

  7. The IT Being says:

    Just found the blog, and really liked it! You are doing great job here, and keep it up like that!

    The IT Being

    <a href="http://gabilgathol.amu.edu.pl/&quot; title="The IT Being">IT News</a>