Innovation Team Tenets - 5
Labs require “T-shaped” people. Broad in their skillsets, but deep in one particular area, e.g. a great all-around developer, but wrote a book on machine learning.
Having T-shaped team members allow you to optimize the ability to resource projects effectively while having the necessary expertise to implement ideas that are truly innovative.
Innovation Team Tenets - 4
To borrow a quote from the classic film Beverly Hills Chihuahua, “We’re Mexican, not Mexican’t.”
People who rush to say why something can’t be done really hurt your culture. People who frame problems in new ways and make original solutions that they are passionate about take your group in whole new directions.
Innovation Team Tenets - 3
This one is a more general management principle, but it’s even more important in a fast-paced environment like a lab.
Be clear about setting expectations up front. If you later aren’t going to be able to meet those expectations, tell me right away and we can reset them. It’s much better to have the hard conversation sooner rather than after we can make any meaningful correction.
Innovation Team Tenets - 2
This Walt Whitman quote sums up the inevitability that given the ambiguity and chaotic nature of labs’ innovation work that senior management will of course communicate contradictions over time.
Do I contradict myself?
Very well then I contradict myself,
(I am large, I contain multitudes.)
Any set of interesting work contains multitudes. A successful team knows to apply disparate strategies for different problems.
As a team member, if you aren’t comfortable with ambiguity and contradictions, you aren’t going to like working in a lab.
Innovation Team Tenets - 1
I recently took a new role within Microsoft leading the development team of Office Labs. This move made me think a lot about the implicit and explicit things a leader can do to create a culture of innovative execution.
In roles where you have been there for a while, over time you’ve come to understand the leader’s unspoken assumptions and principles. In new roles, no one knows those assumptions/principles yet so there is a high probability of people guessing what you meant or not fully understanding your intentions. It turns out a lot of the impedance mismatch arises due to misaligned assumptions and expectations.
This is the first in a series of posts that enumerate the assumptions and principles that I bring as a leader of a lab.
Pivot ideas in HTML
I’ve been playing around with recreating the Pivot interaction model entirely within HTML. A marriage of Quicksand and Lightbox gives a nice, fluid experience for faceted search/browse for a collection with 10s of items.
(note: works best in Chrome; I haven’t gotten around to fixing the script for IE9 changes)
Guitar World Covers Gallery
Getting Past Eye Candy
Hi Brian. I’m the author of that VentureBeat article and the General Manager of Live Labs.
I’m glad that your first reaction is that it “feels” perfect. We’re very proud to have been able to do interesting research and apply it in ways that really do open up new insights in your data in a fun way.
Our Pivot desktop application (the one in the screenshots) at http://www.getpivot.com is our research platform for experimenting with new ideas. As a proof of concept it affords us the ability to release updates more often and stay agile to explore new opportunities as they present themselves. Based on the positive response from early users of that application, we built and recently released an additional way for people to make use of this new type of navigation experience that addresses some key needs from partners and some of the comments in your blog as well. This new release is a Silverlight control called PivotViewer that is free, production quality, and fully supported by the Silverlight product team. This control gives you the full collection viewing and interaction experience within any webpage. Also, because this control is built with Silverlight, you can embed this control in any webpage and get these same kind of experiences cross-platform and cross-browser, including Macs. Seehttp://www.hitched.co.uk/wedding-venues/visual-search.htm for an example.
As for what you call “normal” use cases, I recommend that you use the Pivot desktop application to play around with the Wikipedia collection (located in the collection gallery). The questions I get most are “What if I have millions of items? Do you scale that high?” and “What if my items are textual, not imagery?” The Wikipedia collection shows that the system has the potential to scale gracefully to millions of documents. It also shows that you can take textual documents and create simple visual representations of those documents that retain the essence of the full text document, but also provide new value. The Wikipedia collection is a good example where the concepts still work very well, even though the item is represented by text and simple iconography. We intentionally considered many of the edge cases when we designed the collection experience and our goal has always been to build simple interactions that do scale to virtually any type of content. It’s never going to be perfect for everything, but based on the initial adopters of this technology we’re seeing utility in ways we hadn’t designed for out of the gate particularly in the Business Intelligence community.
I encourage you to download the control from http://silverlight.net/learn/pivotviewer . Build your own collection experiences and let me know where you think the ugly parts are. We’ll work hard to fix them.
In life one must decide whether to conjugate the verb to have or the verb to be.
Moving to the HD web
Seadragon.com gives you an easy mechanism to transform any image on the web into a progressively rendered and zoomable image.
This is a great step towards moving the web away from crappy thumbnails and low res images to full resolution, beautifully rendered images everywhere.
Every image on the web should be this way. Why settle for crappy thumbnails when you get an experience like this for a 461 megapixel image!