Ross Mayfield's Weblog

Google on Collaboration

Notes from a talk at CTC by Matthew Glotzbach from Google Enterprise.

Switching costs for users are nominal in the consumer world.  Enterprise tech is designed for experts by experts.  Vendors with the most checkboxes wins.  Innovation adds to complexity and the net effect is that enterprise technology is becoming less user friendly over time.  Ben Worthen was forwarding his email to his gmail account and has a work and personal identity: "I'm violating our corporate email policy...and I love it!"  Blackberry blurs work and personal life.  Quotes Ray Lane on consumers.  Innovation should not be at the expense of simplicity.  Employees fed up with inability to access information at work, silos of information, business apps require training and expertise and collaboration systems fail at scale. 

He says we are moving from the office worker (1950-1980s) to knowledge worker (1980s-1990s) to self-directed innovator (21st century).  These innovators need talent, direction and information (ideally photographic memory, your company's collective wisdom and the world's information).  They are not process driven, collaborates with a broad network of friends and colleagues, has intermingled personal and work lives, needs information when not at her desk, does not spend mjority of time in a single application and tends not to be patient.

Google's approach to collaboration: Redefining the inbox, where people live.  Email threading is essential at work, a user centric innovation.  Talks through the usual suspects of Google apps.  He lost me near the end.

Mike Rhodin, IBM GM of Lotus, on Mashups

I'm in Boston at the Collaborative Technologies Conference this week.  John Seely Brown kicked it off with a great keynote, but I just got off the red eye and couldn't take notes.  He gave a nice mention of Socialtext and emergent collaboration.

Mike Rhodin, IBM GM of Lotus, gave a talk on Mashups (largely as a rebranding of SoA).  Organizations have been investing in productivity through very formalized systems,but that profit has been maximized, and they now are looking to collaboration for further gain.  Personal productivity in the 1980s focused on standalone use and authoring tools.  Team productivity in the 1990s was driven by LANs with proprietary client/server, doc formats, multi-year development cycles and a focus on email and documents.  Open standards and open source are about to break this open.  Next is what they call the Dynamic Workplace.  Standards, open source, composite applications across boundaries, SoA, etc.  The slide lists lots of buzzwords, but doesn't say what the focus of this era is. 

Sees collaboration at a converging industry, mostly because of disparate user experiences.  An auto company spent 70% looking for information and 30% of the time making decisions, and wanted to reverse the ratio (presumably because they wanted to make decisions based on less information).

Hannover, the next gen of Lotus Notes, provides contextual collaboration with applications dynamically shifting.  Saving time switching apps, tracking revisions.  Hannover will have an updated UI, supports legacy apps, enables composite applications and will be open: ODF provides 125 milion Lotus Notes users alternatives to proprietary Office formats.  Mashups are composite applications that are role-based and in the context as a process.   OMG, he just described Notes as the first Situational Software. 87% of CEOs believe that fundamental change is required in the next two-years.
  Portals will be the initial SoA project for an enterprise.  The primary communication tool in IBM is SameTime (IM), not email.  Presence and location is being integrated into multiple apps and Mike sees things evolving to real-time application solutions. Sametime is build upon Eclipse and can be extended through plugins.  Shows a couple of mashups that leverage backend data and multiple modalities.  There is a heirarchy to to network heirarchy (heterarchy).

By 2009, wikis are predicted to become mainstream collaboration tools in at least half of all companies. Wikis are going to become very very important, disruptive, because they invert the content authoring pyramid (an author gathers sources, out of date, applies wisdoms and publishes something inherently out of date).  A wiki is an inversion, the ideas and framework for the information is contributed and a community can form around that wiki. The reason I think this is going to happen is because it already has happened, in open source communities -- the dynamic is the same.  Content will have a similar path.  If all of your publishing was done through a wiki, what would you use a word processor for?  The value of artifact centric tools will decline.

He concludes by highlighting tagging and activity-centric collaboration (a collection of materials, communications and processes that emerge when peopel work together on a common goal).

Establishing Common Fact in Fragmented Media

Jim Griffin provided great food for thought at Aula with the insight that mainstream media used to provide a basis of common facts for society to hold conversations of value and policy upon.  Regardless of they were indeed facts, it differs from fragmented media where each (e.g. Fox News) frames their own facts, leaving society to debate their differing facts (e.g. was 9/11 ordered by Saddam?) instead of higher value conversations that foster community.  And as advertising drives media towards isolated personalization, Jim's insight is that increased media choice when the network is a delivery mechanism, particularly in the mainstream, has a different societal impact than when the network is an interaction mechanism.  With the former, communities will erode, providing the latter a call to action for social media to build anew.

The first way to counter this trend is familiar to those of us engaged in social media -- leverage participation to strengthen existing communities and foster new ones.  The concern here is while early adopters have developed practices and tools to be more broadly connected, echo chamber arguments aside, mainstream crossover is a challenge.  Fifty million members in MySpace is more of a network than a community with a common grounding of media facts.  Stronger tie networks like Meetups or core communities engaged in collaborative intelligence hold promise.

The second way to attack this problem is the underlying advertising model.  Hyper-targeted advertising, notably keyword targeting, assumes intention to buy within search, but ignores social influence on transactions.  Alternative ad models that incorporate multiple media influences and the participation of social networks on individual buying decisions effect the shape of media.  Composite form follows composite funding.  But these forms have yet to arise.

The third and most promising dynamic that runs counter to fragmentation is how social media is connected to itself.  That is, unlike broadcast, media influences other media and the primary user activity is sharing.  When a fact reaches escape velocity, memetics serve to establish common fact. A false mainstream fact is checked by the blogosphere.  A fact gains credence from it's distribution.

In this post I'm not using the word fact in absolute terms, as I don't believe in absolute truth and would instead offer that facts are social.  The litmus test for if a common fact can be established across fragmented media could be simple: if a left-wing blog can cultivate a meme to the point of leverage where Fox News broadcasts it.  In that case, if its on Fox, it's got to be true.

The reason common fact matters is the accelerating pace of change presents a grave risk if we cannot collectively make sense of it.  In a world of decentralized and fragmented media, is there a source of collective fact-oriented media and underlying community.  I think there is one, so far -- Wikipedia.

Aula

I'm at Aula, a meeting of creative minds, in Helsinki.  Always wonderful to be in Helsinki, especially in the summer.  Here are some bits and pieces, an open note pad (but note I didn't blog the best sessions from folks like Clay Shirky, Joi Ito, Dan Gillmor, danah boyd, Alastair Curtis, Martin Varsavsky and others):

UPDATE: Teemu has good notes from the public session.

Jim Griffin gave an impassioned talk about Me Media.  His view of media fragmentation is that it is destructive towards communities.  Finland happens to be the last bastion, where most people get their news from the same TV channels.  While generally in the blogosphere we decry such Cronkite culture, We Media provides a basis of facts that enable people to debate value or policy.  With Me Media fragmentation we all get our own Cronkites, preferred by advertisers, we argue about facts.  Jim's hope is that we use social media to help defend existing and create new communities.  Jim uses language counter to Dan Gillmor's definition of We Media, but they share the same concerns.

Un-wiki: gets quotes from Wikipedia the free encyclopedia's Deletion Log  The Deletion Log is a list of all the pages that have abused Wiki's democratic remit; it is the last stop on the way to destruction.
It has small amounts of the offending texts. I'd like to see a wiki comprised of deleted pages.

While on the topic of wikis, David Berlind has a thoughtful and researched post on wikiCalc.

Instructables: Going through time linearly means there is always a sequence of steps for documentation.

Adam Greenfield has a new book on ubiquitous computing called Everywhere. C. Weiser from Xerox PARC: the mainframe is powerful and central instrument that supports community.  From sharing one computational resource, to a future where computation will "be invisible, but everywhere around us."  Ubiquitous computing is embedded, wireless, imperceptible, multiple and post-gui.  Dissolves in behavior.  A technological substrate that tends to colonize everyday life, such as the Panopticon or an internet toilet that analyzes your waste and transmits to your doctor.  This changes the presentation of self.  Alan Finder: for some online persona disrupts a resume.  We don't choose to interact with ubiquitous computing, sensors engage you.  Proposes five ethical principles:

1. default to harmlessness
2. be self-disclosing
3. be conservative of face
4. be conservative of time
5. be deniable

Saul Griffith: Computation is 1-2% of energy cost, this will substantially increase with Ubicomp.  Cory suggests we should have the right to falsify our RFID data.

Some personal reflections... the larger pattern between the media and fab movements seems to be that with both media and computation we used to have central constructs that built community around it, but now may work against it.  I liked how Adam began his talk with mainframe as community, but I would have liked to explore Ubicomp as community further.  For example, right now I am sharing a computing resource, the wifi and limited connectivity off of a tiny little island with 40 people in a round room.  We are sharing, but there is little coordination on how to use this resource. Recognizing as I write this, I turn off Skype to stop sharing in-room resources with my extended network.  Having a resource be an organizing construct for a community is of course, a model of scarcity.  It wasn't just consumption as time sharing that fostered community, but what they built together, the abundance of their ideas and social capital.  I'm interested in how Ubicomp could at the least help us share resources when scarce, and at the most, discover our ambient abundance.

Joshua Ramo: works for Henry Kissenger and travels often at high speeds. Unlike understanding, Knowledge is falsifyable.  It's possible to understand things without knowing them.  Studying enlightenment means studying how we move from knowledge to understanding. Kissenger's internet ignorancy theory: you used to have to learn a lot to understand things, but now knowledge is easier to get, but harder to get understanding. Invading Iraq was a Google-compliant decision.  Budda reached enlightenment at 36, at an annual average velocity of 0 miles an hour, mostly sitting under one tree.  The new speed of enlightenment must be greater than 0.  Velocity is not a guarantee of enlightenment. Dominant feature of the world today is constant and accelerating change.  If you want to be enlightened you  This leads to a corollary observation that modernitys greatest charm and greatest terror is the unpredictability associated with ceaseless innovaiton No one 100 years ago predicted the web or Al-qaeda. A rlife of rapid change is good because it puts you in harmony with the fundamental truths of the world.  Side effects: hard to hold on to emotional attachments, have shocking moments of intimacy.  The kind of enlightement you get at 0 or high speed: the faster you go, the more you have to be internally stabliizing.  Modern life is a ceaseless earthwuake; rapid velocity is the best way to train your own stabilizing mechanism.

Matt Jones  brought along a backup Matt just in case.  There are a lot of Matts on the internet.    Cities and digits are the two megatrends.  Parkouriustes, the French notion of moving through cities with mad skillz:  The most important element is the harmony between you and the obstacle; the movement has to be elegant... If you manage to pass over the fence elegantly—that's beautiful, rather than saying ‘I jumped the lot.' What's the point in that? Cities are our new nature.  Practical superhumanity.  Matt Webb talks about supersenses: using information is nothing like moving through space, yet we persist with using direct navigational metaphors.  Quinn embedded magnets into her fingertips, and now starts to gain a different sense of information flow, stickiness and texture from metals.  Rhinos see the world through smell.  How to create a robot-readable planet.  Parkours doing a tour of London had to feel out the difference of the streets with their feet compared to Paris.  Why only create eye candy when we can create foot candy?  Buckminster Fuller: Make the world work for 100% of humanity, in the shortest possible time, through spontaneous cooperation, without ecological offence or the disadvantage of anyone.

At lunch someone showed me the new Nokia E61 device that looks like a Blackberry-killer.

WidSets: wireless widgets

Jyri Engeström and Mika Raento on social peripheral vision.  Phones are designed with the assumption that you know who you want to call before you do.  You need to process social signals before using the device.  Jaiku, their startup is looking to augment basic functions of a phone by pasting onto it what is happening on the internet.  If you can't find anyone in your contact book, you can search a directory made of everyone's contacts.  Calendars let you share future events to let you plan together.  The demo shows very rich profiles based on phone usage (automatic data) and more social signals (more manual) -- which provides a different form of Presence.  In usage, people still call regardless of presence, but when someone doesn't answer, you leverage the presence to understand why.  Integrated IM is more convenient than SMS, and includes group messaging.

Justin Hall plays a lot of video games and surfs the web a lot, today he is talking about Bud.com.  Games should understand you and adapt to you.  Justin is a one man fishing machine on WoW.  Oblivion is a massively single player online game, all the scope of WoW and none of the socialization.  It watches the way you walk around the world and increases stats depending on how you move.  Active Timer is my-ware to do your own data mining.  What would you learn about me from how I use the web?  Justin is prototyping a system that profiles you, adapts the UI, rewards you (congrats! you leveled in using email!) to create Passively Multiplayer Online Gaming.

Habbo Hotel is a browser based virtual world for teenagers.  Now they are prototyping "Mini Habbo" for mobile use.  Prototype one focused on the UI, prototype 2 focused on network performance.  The demo is really cool, you could see playing with that thing in your pocket.  Commercial challenges: teens dont have fancy phones, data traffic costs, missing internet access settings and use context aspects of mobile virtual worlds are still unclear.  There is room for exploration: potential for new products and relatively easy to develop. 

Similarly, Stardoll is a site for teenage girls, where they design toys, not games, with 100k visitors a day.

Alice Taylor talks about gaming and broadcasting.  BBC is transitioning from broadcast to digital media.  Broadcast is playout, games are playin.  In-world broadcasting is delivery of existing material, such as listening to the real radio while playing a car racing game, or showing TV on a big screen in Second Life.  Environments recognize that WoW is an established environment...so now the movie is in the pipeline, film what they are doing within the game.  In your average American household 28 hours of TV is watched per week -- which drops to 6 hours per week when there are games (played an average of 24).

Berlin

I'm in Berlin for a customer visit while on my way to Aula '06 in Helsinki.  I had hoped to suprise Felix from Plazes by being in a Plaze near to him.  But then his presence showed Westin Palo Alto and mine Westin Berlin.

My little bit of World Cup fever has been watching games in the Fan Fest area of Brandenberg gate.  Staying in the same hotel as the Croatia team, who plays Brazil today.  They'll probably be crushed like the USA was yesterday.  While my customer is generally from a buttoned-up German company, today we plan on meeting in jerseys and shorts.

Saturday I wore shorts to a panel on community at the eBayDevCon in Vegas.  Got some great feedback and interest on wikiCalc. In that case my shorts were hidden behind a podium. Not a bad way to do business during the summer. 

wikiCalc Screencast

Dan Bricklin prepared a screencast (7Mb, .swf) as an introduction to using wikiCalc.  In it, you will see him paste from Excel and mashup stock quote and charting feeds.  The Associated Press posted a in-depth profile of Dan that's a great read.  Also read his blog posts working with Socialtext and the Beta release.

UPDATE: Podcast with Eric Lundquist from eWeek and Dan Bricklin.

Socialtext Partners with Dan Bricklin on wikiCalc

Today Socialtext announced a partnership with Dan Bricklin (inventor of VisiCalc) to exclusively distribute, redistribute and co-develop wikiCalc -- the social spreadsheet. Dan brings a rich understanding not just of spreadsheets, but also open source and social software. You can thank VisiCalc, the original killer app, for the Personal Computing revolution and bringing PCs into enterprises from the bottom up. It may take some time, but Socialtext, wikiCalc and the community will develop an important contribution to Social Computing. I'd like to share how this came together, what it means and where we are going.

David Weinberger re-introduced me to Dan, whom I met at a couple of conferences. The deal was a calculated decision that took almost six months. We needed to build a shared understanding for how we can work together and work with a community. We share common principles in open source and designing social software. We see an opportunity to change the way people work together, that is different from an office on the web.

Andy McAfee defined Enterprise 2.0 as the use of freeform social software within companies: freeform in that the application does not impose structure prior to use. Since the inception of Socialtext, we have avoided the temptations of structure, not just because one man's structure is another's barrier, but because it immediately divides the world into those who can structure it, and those who cannot.

Wikis begin as a blank page, just like spreadsheets. Some think this is a weakness, but it is actually a strength -- because it asks all the right questions. How should we use this tool? What should we apply this to? What kind of buckets should we put information in? What's my role? To make a wiki work, and all IT for that matter, agreement on how to use the tool makes it work.

When Ward Cunningham invented the wiki ten years ago, one of his core design principles was a collaboration environment for experts and novices alike. This equality of usability matters, especially in the context of an enterprise. Some approaches assume that users will learn a new proprietary scripting language to create or modify structure, but programming will never be a novice activity. Writing and linking can be done by novices, and so too using a spreadsheet. Spreadsheets are the most widely adopted interface for creating and modifying structure, databases and applications.

One way that a wiki, and wikiCalc, enables social use of a tool without imposing structure is the properties of transparency and memory. In a wiki, a copy of every edit is saved with attribution. This supports trust and accountability. Given the nature of spreadsheets, wikiCalc takes this one step further, creating a revision history of every single edit to any cell -- not just when you save and the undos before. This not only holds the promise of tracking down errors, but provides an audit trail.

But spreadsheets, like other killer PC apps, were not designed for a networked world, but for a single users. Today the problem isn't just people playing email volleyball all day digging for who did what to each revision. Work is social, information can be linked and data comes in feeds. Today it isn't the problems of productivity kept personal, but the opportunity to build computing that is social.

Dan just released wikiCalc Beta as an Open Source GPL distribution. Right now you can use and modify a web-based spreadsheet of your own. We hope you will join this community. In the coming days we will release wikiCalc under a more liberal and commercial-friendly distribution. I'll let you determine the differentiating features, but the primary one isn't in the code -- it's that it is Open Source. We see great promise for an enterprise-friendly wikiCalc.

You may know Socialtext's commitment to Open Source has been a foundation of our company values. Today I can share that the Socialtext Open Source Edition, functionally equivalent to our commercial wikis for users, will be released next month at OSCON. Yes, I know it's a little odd for startups to communicate such news in advance, but there is nothing to fear, and it's part of being open with a community. Wikiwyg, wikiCalc and the Socialtext Open Source Edition will all be released under the same OSI complaint Open Source license, based upon the Mozilla Public License. For more information on Open Source licensing, see Dan's great video.

It's a delight to work with Dan. We have a lot of work ahead of us and hope you will join us.

UPDATE: The Associated Press has an in-depth profile of Dan.

6/6/6

I'll explain later.

CEO's Guide to Enterprise 2.0

BusinessWeek's Rob Hof on a CEO's guide to Enterprise 2.0:

For all its appeal to the young and the wired, Web 2.0 may end up making its greatest impact in business. And that could usher in more changes in corporations, already in the throes of such tech-driven transformations as globalization and outsourcing. Indeed, what some are calling Enterprise 2.0 could flatten a raft of organizational boundaries -- between managers and employees and between the company and its partners and customers. Says Don Tapscott, CEO of the Toronto tech think tank New Paradigm and co-author of The Naked Corporation: "It's the biggest change in the organization of the corporation in a century."

Rob lays out the guide and tipsheet, with qualifiers, on his blog:

As a "CEO's Guide to Technology," it's clearly aimed at executives, so those of you who know this stuff cold may not be surprised. But we in Silicon Valley tend to forget that most of the rest of the world hasn't even heard of Web 2.0. In a podcast, Tim O'Reilly explains the basics, and provides a little more insight into the flap over his Web 2.0 Conference partner taking out a controversial service mark on the term as applied to conferences. And in a Q&A, Ray Lane provides the VC's view.

When he interviewed me, one thread was how things have changed over the last three years in the wiki business.  The broader cultural trend for consumer adoption of these tools has, at the least, made it easier to introduce them to the enterprise.  Less to explain or to train.  Every now and then a CEO catches the wiki bug after discovering Wikipedia on her own.

Nonetheless, the notions behind Web 2.0 clearly hold great potential for businesses -- and peril for those that ignore them. Potentially, these Web 2.0 services could help solve some vexing problems for corporations that current software and online services have yet to tackle.

For one, companies are struggling to overcome problems with current online communications, whether it's e-mail spam or the costs of maintaining company intranets that few employees use. So they're now starting to experiment with a growing array of collaborative services, such as wikis. Says Ross Mayfield, CEO of the corporate wiki firm Socialtext: "Now, most everybody I talk to knows what Wikipedia is -- and it's not a stretch for them to imagine a company Wikipedia."

MORE FLEXIBLE. And not just imagine -- Dresdner Kleinwort Wasserstein, for instance, uses a Socialtext wiki instead of e-mail to create meeting agendas and post training videos for new hires. Six months after launching it, traffic on the 2,000-page wiki, used by a quarter of the bank's workforce, already has surpassed that of the company's intranet (see BW Online, 11/24/05, "E-Mail Is So Five Minutes Ago").

Corporations also are balking at installing big, multimillion dollar software programs that can take years to roll out -- and then aren't flexible enough to adapt to new business needs. "They're clunky and awkward and don't encourage participation," grumbles Dion Hinchcliffe, chief technology officer of Washington, D.C. tech consultant Sphere of Influence...

I think what CEOs get about Enterprise 2.0 may surprise you.

• They are proponents of transparency.  Not just because it is the underlying objective of SarBox, but they couldn't get to their position through hoarding and recognize that six levels of reporting tends to water down the truth they need to make decisions.
• They understand that adoption matters.  Because at their level, it's a major input to ROI and want to get what they are paying for.
• With suprising frequency, they use RSS newsreaders.  This is anecdotal, but otherwise a good part of a CEO's role is reading.
• Control is less of an issue than for middle managers.  Again, to advance, they have had to lead more than manage, and delegate.

The Ray Lane interview provides a similar view:

Enterprises are fine with them. I've talked to a lot of chief information officers about this. As a group, financial institutions are wary because of regulations. They can't even use instant messaging without logging and archiving them. They've got to have a record of everything.

But other than that group, every other kind of industry and CIO I've talked to absolutely buys into it and says, "Bring it on." Obviously, you have to meet security concerns. But I don't find chief executives wary of podcasts, blogs, wikis, or social networks.

The point being, CEOs are ready for the shift, but need to work with vendors who have adapted social software for the enterprise within security requirements.

eBay Wiki and Blogs

eBay is perhaps the greatest online community, but built on 1.0 tools.  Now the news is eBay launching blogs, wikis and search tags next weekend.  I love this acronymism from the wiki help page:

What

I

Know

Is

I'm speaking on a panel on Web 2.0: Tools for Creating a Community with Jeff Clavier, Laura Merling, Kevin Rose at the eBay Developer conference.  Lots to talk about.  And talk has always been (de)central for eBay, as I noted when they bought Skype.

Pierre Omidyar once explained to me that one of the smartest things he did when starting eBay was to not constrain communication around his market -- by publishing email addresses.  He was suggesting to me that we open the Socialtext Customer Exchange, but the core insight is more valuable.  Back when I was running a B2B exchange, this was considered a contrarian move.  After all, it let buyers and sellers circumvent your transaction fees in some cases.  But letting go of control fosters liquidity.  Especially when you couldn't possibly structure communications to fit all transactions.  Today I would venture that most of the communication on eBay's transactions are out-of-band. Other communities with emergent liquidity such as Craigslist succeed by enabling even further out-of-band communication.

But there are some unique challenges for eBay's next generation community.  Principally, while markets are conversations, social software gets spammed.  And while eBay desired a fact-oriented wiki, it both inherets community values and doesn't have the luxury of cultivating it over time once the wiki is made public to the masses.  I've chatted with eBay about blogs and wikis since the Omidyar Network investment, and these folks get community.  But the social contract seems borrowed from Wikipedia and it will be interesting to see it be rewritten.