Enterprise 2.0: Sam Weber

Last day at the conference — half-day, really — and the first breakout session is with Sam Weber of KnowNow, discussing “RSS: Bridging the Gap Between the People and the Information that Drive Business”. I’ve been looking at ways to bring RSS into the enterprise for quite a while, mostly by nagging the BPM vendors to include it in their products.

Weber starts by saying that they are now completely an RSS company (actually, they also support ATOM), and previously did system-to-system integration using techniques such as REST: essentially, lightweight SOA. It makes perfect sense to me that moving to a focus on RSS is a logical progression from that. They make an RSS reader and a desktop alert tool, and their server is available as a hosted or on-site solution.

In a case study, he discusses a company of 30,000+ employees that has operations in the US and India, with 30 intranets, portals and knowledge bases, and thousands of internal blogs. Their challenge was internal communications across their employee base, and by using RSS for information distribution they were able to deliver consistent communications in a timely manner, without the information being lost in the employee’s email boxes. In other words, as I’m starting to see, RSS is becoming a sort of high priority email replacement, and to avoid clogging up internal email systems with broadcast emails.

He restated his main points from his 6-minute Launch Pad presentation yesterday: email is overused, with over 50% being junk; static portals are useful to less than 20% of users since they require too much action on the part of the user to find the information that they required; search has only about a 50% success rate and also relies on the user to have the skills required to find what they want. In fact, an IDC study showed that the average employee spends 9-10 hours per weeks searching for information, and is only successful half the time (it sounds like a lot more people need a tool like Google Desktop Search, or X1, which I use all the time), and that a 1000-person company wastes $48k per week on the inability to find information.

The solution, according to Weber, is RSS in the enterprise, or what they call live information management. To do this:

  1. Access and monitor information sources, rather than replicating data for the purposes of syndication; this includes monitoring information sources that don’t have RSS feeds.
  2. Automate relevancy, by leveraging this “newly freed content” to allow the system to try to determine what is relevant to a particular user and push them that information.
  3. I blinked, and missed step 3. 🙂 I think that it was more about pushing the information, that is, transform and deliver the data.
  4. Capture user behaviour so as to refine the relevancy algorithms.

With all of this, there needs to be some core system capabilities such as aggregation (to avoid unnecessary duplication), filtering to meet relevancy requirements, data transformation to allow access to sources that don’t have RSS feeds, security, alerts and notifications, and content-based routing of the feeds to those who need them. After that, the information needs to be delivered to the end users, which is done via RSS.

Weber sees the following minimum requirements for a live information management system: monitor all sources inside and outside the enterprise; match content to users based on relevancy; leverage the network effect; deliver information to users as available; provide enterprise security and management; and enable end-user personalization and control. This seems to be a bit redundant with the last two slides, but is a good summary.

His next case study is Wells Fargo, which uses Teradata for collecting information in six major data warehouses; their challenges were around synchronizing metadata and ensuring that data and schema errors were corrected quickly. They’re now using KnowNow (in a customized solution) to monitor parts of the ETL process related to a risk dataset, and push notifications of any problems to the appropriate people via RSS.

Next, he discussed another banking case study (unnamed) where the central organization was sending out more than 80 internal newsletters and alerts each week to all employees; when this bogged down their systems, they moved it to SharePoint, where people actually found that it took more time to find things than in their old Lotus Notes email system. Now, they’re moving to an RSS model for distribution, and find themselves saving $750k per year.

He finished up with options for the enterprise to adopting RSS:

  • Wait, and just use RSS in the systems that offer it today; the big vendors will likely take 2-5 years to bring in RSS formatting.
  • Continue to (over)use email as the main collaboration tool, and bet on portals and search technologies.
  • Implement an enterprise syndication solution, which leverages the information that you have now and allows information to be selectively pushed to people, or allow them to select their own subscriptions.

I have no idea how well KnowNow’s product works, but the concepts that Weber discussed really hit home for me and many of my clients. There’s a Forrester white paper about Enterprise RSS available on the KnowNow site if you want to learn more.

Enterprise 2.0: Hot Videos

No, not that kind of video — these are hot-off-the-press videos of the keynote sessions at the Enterprise 2.0 conference. I stopped by the Altus display at the trade show after the sessions completed today and saw a bit more about the technology behind what’s serving these up. First of all, less than 24 hours after a live session, they have the video up here, complete with a fully searchable transcript, the “talking head”, and the slides, all synchronized. If that weren’t hot enough, note that there’s an RSS feed icon on the page; click on that, copy the URL for the vodcasts, paste it into iTunes, and you’ll be watching the slides with audio on your video iPod.

Enterprise 2.0: Tapscott Panel

Following on directly from his talk, Don Tapscott moderated a panel including Ross Mayfield of SocialText, Kim Polese of SpikeSource (who I saw speak about open source at the CMU West New Software Industry conference in April) and Joe Schueller of Proctor & Gamble.

Mayfield started off with some ideas about the complexity of social networks, and how some vendors attempt to translate that complexity into the software rather than providing simpler tools that encourage emergent behaviour by leaving the complexity with the people part of it. He illustrated the power law of participation, making the distinction between collective intelligence and collaborative intelligence, before moving on to a discussion of what to wiki. First, focussed on people, you can do an internal equivalent to Wikipedia, such as a part of your intranet; the challenge is that there may not be a specific goal associated with this. Second, focus on practice, and create a global glossary or how-tos which provides a practical but non-political reference application. Third, focus on project by creating a forum for project communication and lightweight content management. Fourth, focus on process by using a wiki for handling process exceptions. He finished up by announcing that SocialText launched WikiWidgets today, although I’m not clear at this point what exactly that is.

Polese talked about the enterprise benefits of Web 2.0, but also covered off some of the concerns: cost of the administration of point solutions, security, importing/including data from current information sources, and enterprise content management. She discussed their products, some of which provide testing and support frameworks for using open source in the enterprise, and one which appears to be an enterprise collaboration platform. I’ve seen so many things called “enterprise collaboration platforms” in the last two days, this is starting to be a bit meaningless.

Schueller was up last, discussing how P&G has relabelled their product R&D department to C&D, for Connect & Develop. They see the advantage of collaborating not just to harness technical ideas from all over, but cultural issues that might be relevant to a consumer products company like P&G. He sees this as a key part of business agility, as well as reducing product development costs and just creating better products from better ideas. They’re focussing on their core competencies rather than having to do everything themselves, which is something that’s happened in many industries such as manufacturing, but the idea that you can outsource your ideas is pretty innovative.

It’s like a sauna in the conference room and we’re now at 4:30 on a day that started with 8am sessions; and as we tail off to general questions on the panel, people start to trickle out of the room.

There was a good discussion on the potential for replacing email within the enterprise; Schueller said that he once made a bold statement that they’d just get rid of it, then had to eat those words as he discovered the (current) futility of declaring war on email. I think that there’s great opportunities to reduce email, using things like wikis for collaboration, and RSS feeds for one-way information dissemination (as we saw in the Bank of America presentation this morning), but we’re not going to see a significant decrease in email until more of the boomers retire and more of the MySpace generation gets into corporations to help shift the culture.

Enterprise 2.0: Don Tapscott

Don Tapscott talked about the concepts from his book Wikinomics, which I finally got around to reading recently (okay, I confess, I didn’t read the whole thing…). He states that the web is becoming the new mode of production, and is facilitating the move from the old closed hierarchical corporation to the open networked enterprise. He sees four drivers for change:

  • Web 2.0, which changes the way that we interact with computers.
  • The net generation, who grew up digital and also are much more used to multi-tasking across multiple channels: reading, watching TV and surfing the internet at the same time. They see email as a “formal technology”, suitable for sending a thank you letter to one of your parents’ friends, for example.
  • The social revolution, characterized by the rise of sites like Facebook.
  • The economic revolution, including new business models to take advantage of mass collaboration.

He covered his four major principles of being open (as an organization): peering, being open, sharing, and acting global. I did like his quote about being open: “if you’re going to be naked, you’d better be buff”, namely, you probably want to have your company in pretty respectable shape before exposing it to your business partners and customers.

He closed with the new business models that he discussed in the book: peer pioneers (e.g., Spikesource, who is up speaking later this afternoon); ideagoras (open idea markets, e.g., Proctor & Gamble, who will have half of their innovations come from outside of the corporation by the end of this year); prosumers (where customers become producers, e.g., Second Life); the new Alexandrians (e.g., Google Earth Avian Flu mashup); open platforms (e.g., Amazon API); the global plant floor (e.g., Boeing creating a peer-produced aircraft with their suppliers); and the wiki workplace (e.g. Geek Squad product management within Best Buy).

In direct relation to something said in the collective intelligence session this morning, Tapscott said “don’t fear theft of your intellectual property, fear obscurity.”

Apparently this presentation is pretty well-worn: Martin Cleaver, who’s sitting next to me, has heard it three times, and Mark Kuznicki, who’s chatting with us on a group Skype chat from Toronto (I think) says that he can recite it by heart; in fact, he described it as “peer production, blah blah blah, user-generated content, blah blah blah, crowdsourcing, blah blah blah”. This the first time that I’ve heard it, and it’s pretty good the first time around, but Tapscott may want to get himself some new material to keep the repeat attendees happy.

Enterprise 2.0: From the Labs

David Coleman is hosting a panel that includes Bob McCandless and Chad Ata of BrightCom, Irene Greif of IBM, and Denis Browne of SAP of what’s happening in their experimental environments right now, which may not be anything that’s going to become a product. Each has 10 minutes, so this is more like a standard panel than the speed-dating vibe of the previous Launch Pad session.

McCandless started out talking about BrightCom, which makes telepresence products, and some of the work that they’re doing in gaze correction and perspective-corrected viewing so that it appears that the people on the other side of the conference are both at the table with you and looking at you. He’s talking about immersive telepresence, where you’re in a fully rendered environment that appears truly seamless since you appear as a photo-realistic avatar rather than using a literal video of you, and how new display technologies are going to have to improve in order to support that. He’s demonstrating the ideas by using Second Life, where he’s popped his telepresence avatar directly in instead of his normal Second Life avatar. He claims that this has the potential to become “indistinguishable from reality”, and that immersive technology is the next generation after telepresence.

Greif talked about IBM’s Many Eyes, a social data analysis site. She started by showing the Baby Name Wizard data visualization, which somehow reminded me of the Hans Rosling TEDtalk and his company GapMinder; as she got into Many Eyes itself, it’s very much like GapMinder. The vision, however, is to use tools like this for enterprise data, including that which is residing on people’s hard drives.

Browne talked about what SAP is doing with enterprise widgets. He makes a great point that IT departments are still in reaction mode with respect to Web 2.0 applications, and that in many cases, the business side is bringing them in. He sees three ways to implement Enterprise 2.0: extend existing applications, build new applications, and facilitate enterprise mashups. He also makes an analogy of Enterprise 2.0 solutions to iTunes, where they need to provide end user experience, distribution, security and development tools. He showed the SAP widgets, which is a Google gadget-like toolbar of widgets, such as drilldowns into sales prospecting data.

Enterprise 2.0: Launch Pad

Michael Sampson of Collaboration Success Advisors chaired Launch Pad, where four companies each had six minutes to talk about a new product that they’re launching. David Coleman and Stowe Boyd each provided a minute of critique, then at the end we voted via SMS for our favourite.

Franco Dal Molin of Collanos: team collaboration across the corporate firewall, so that it can include outsourcers and contractors. They have a synchronized peer-to-peer workspace solution which is server-less, allowing for sharing information both real-time and asynchronously. You invite others to your workspace, and what you’re doing synchs with what they’re doing. It can include any types of content or file types, but also structure content such as tasks and links. There’s versioning of the content. Each project is presented in its own tab for organization, and presumably you can share different different tabs with different people. Workplace was released about six months ago, and today, they’re announcing a voice calling add-on to allow for voice and video calls, including voice mail and call forwarding. He invited us to join the Collanosphere. [I’m wondering if you’re just looking at the product, is that a Collanoscopy?]

One of Stowe Boyd’s comments was that the problem with P2P is that if you’re never online at the same time as your collaborators, then you never synch with them; I see that in the Skype group chat that I belong to, where the chat history is actually maintained through P2P synchronization and I’ll suddenly get a big chunk of missing chat history come flooding in when someone comes online.

Avinoam Nowogrodski from Clarizen, who I’ll be interviewing later today: on-demand, collaborative project management. He had a lot of great marketing graphics that talked about collaborative project networks, but I’m not left with much of an overall impression of what their software actually does (aside from the fact that it’s called “project management”, so I assume that it does some sort of, um, project management).

Dave Peak of LiquidTalk: some of the challenges in business today are that an increasingly remote workforce limits collaboration; the Web 2.0 generation enters the workforce and have higher expectations; and roles are getting tougher, leading to higher turnover in employees. The result is a disconnected, disengaged workforce, and LiquidTalk counters this with mobile workforce engagement: create, find, organize and push audio/video business content to mobile devices. Personally, I’m not sure that I want to consume a lot of information via voice; although I listen to podcasts a lot, if I’m just trying to extract some key data, text is way easier.

Sam Weber of KnowNow was up last: his title slide claims that they’re the Live Information Management Application for Today’s Enterprise. He states that email is overused (and has become a storage device rather than a communication platform), static portals are broken, and search isn’t the answer. They monitor a wide variety of information sources — blogs, wikis, web servers, email, portals and more — and aggregate it, apply filtering, then push it out the critical information via RSS to create channels in a user portal environment similar to Netvibes. I agree with the critics that aggregation and RSS feeds are the way to go for information consumption in the future, so this one may have some legs.

The audience vote on which of these has the game-changing proposition: KnowNow.

Enterprise 2.0: BEA

After watching Ajay Gandhi in the panel just before lunch, I meet up with him for an update on what’s been happening with their Enterprise 2.0 products since BEA Participate last month. It looks like they’ll move out of beta to general release some time in July, and they’re starting to line up customers for the GA products from around the world, and across telecom, public sector and a variety of other vertical industries.

We talked about what beta customers are doing with Pages, their mashup platform, and there’s a number of things that I didn’t expect. First of all, customers are using it as a wiki platform for team collaboration: although not as wiki-functional as a true wiki platform, it can easily integrate in other content and data sources, which a wiki platform typically can’t do. Other customers are using it as a lightweight portal builder (in my opinion, it looks like it could replace WebLogic Portal for some simpler applications, although that’s not BEA’s positioning), and also using it as a tool to create portlets that can be consumed by the WebLogic Portal environment. Since Pages allows you to create a portlet directly from sources as diverse as REST or an RSS feed, that adds a lot of value.

We also discussed a bit of the roadmap for future versions, the most exciting of which (for me) is the integration with AquaLogic BPM to both invoke and monitor processes. They’re also going to strengthen their blog and wiki capabilities, and improve the tooling, and he mentioned that they’re considering a hosted version as well.

More and more, I’m seeing mashups — especially via enterprise-focussed platforms like Pages — to be the future of what is now called end-user computing: the small but necessary applications that are built by semi-technical people that exist within business units, created in the absence of a sufficiently agile IT development group. I’ll definitely be looking more at this theme in the coming months, especially at Mashup Camp next month.

Enterprise 2.0: Enterprise 2.0 Mashups

David Coleman of Collaborative Strategies ran a panel on Enterprise Mashups that included Ajay Gandhi of BEA, Rod Smith of IBM, Eric Hoffert of ShareMethods (who I had a great chat with yesterday at a break) and Lee Buck of Near-Time.

Coleman defines a mashup as something that can be working in less than a day, and declared that anything that takes longer than that is an integration project. I’m not sure I agree with that strict time definition, but it’s definitely in ballpark. He started out with some interesting slides defining mashups, since it’s likely that many in the audience aren’t familiar with mashups. His talk would have been a bit more impactful if he had taken the Bluetooth headset out of his ear, however…

Gandhi was up next to cover BEA’s mashup environment (he used the term “adhocracies” to describe some of the types of collaboration that are going on, which I love), and discussed how enterprise mashups are different than consumer internet mashups since they have to plug in to corporate data and provide some degree of security and control. He primarily covered AquaLogic Pages, which I saw at the recent BEA user conference. The nice thing about this panel is that each participant is expected to not just present, but demo, so he moved over to a Pages demo, although he didn’t do a live inclusion of my blog during the demo as we had from the stage at BEA Participate. He showed how to create a new data space, or page, which includes feeds from other sources, interactively-entered text, and other content; each page automatically has its own RSS feed. He then created a simple web application in Pages by combining a data object containing names and addresses of potential sales prospects with a map widget, where the map and data widgets interact with the map detecting any address information in the data and plotting the prospects on the map.

Hoffert gave a quick presentation on mashups as a model for building and selling applications, where revenue can be shared across the mashup participants, and the need for better standards to allow mashup components to interact. They’re involved in Open Simple Ajax Mashup (OpenSAM), which is intended to more easily allow for multi-component, multi-vendor enterprise mashups — this is another twist on the same issue being addressed by at least one other vendor by considering a sort of mashup ESB. Hoffert then demonstrated ShareOffice, one of their products, which is a mashup built on five different components for creating, editing and sharing documents amongst members of a sales team, and includes SalesForce.com data as well as document authoring and management.

We moved over to Smith, who started with some discussion of why companies do mashups, which is usually about how to reuse assets in ways that they were not intended. The emerging mashup ecosystem has to allow ease of access, have data and components be designed for re-mixability, and be ready in a corporate culture sense for the emergent behaviour and applications that will result. He demonstrated QEDwiki, which I’ve seen at Mashup Camp last year and in a few other presentations since then, although I haven’t tried it out yet. The page creation environment allows you to drag and drop components from a palette to the page, then open up properties for each component in order to create the links between them. The application that he showed, which was built in 17 hours, so fits the “less than one day” criteria if you’re a typical developer, combines two views of data about shipping vessels (a list, then a drill-down on the selected vessel) with a map of the location and relevant weather data. One of their widgets allows you to create a feed from a SQL statement, or from an Excel/CSV file, in order to integrate into their environment; this further supports my thoughts of (RSS) feeds as being the next great thing in what used to be called end-user computing. You can’t use third-party widgets, such as Google gadgets, although Smith sees that having widget interface standards is going to be essential as we move forward: essentially the same message that we heard from Hoffert about OpenSAM. It appears that this will all be part of Info 2.0, which we heard about yesterday in IBM’s Enterprise 2.0-o-rama presentation.

Last up was Buck from Near-Time, which looks at cross-organizational collaboration. He discussed “people mashups”, namely online collaboration outside corporate boundaries, where there is no single authority. They use methods such as an OpenID gateway to an internal LDAP directly as a way to help facilitate that collaboration, and provide tools for secure content sharing. Sorry guy, this is not a mashup in the sense that anyone attending this panel means, although it might be an interesting collaboration environment.

An audience member asked about process mashups rather than data mashups; Smith responded with an example about Tivoli information being available for exposure in a mashup environment, which is not really what I was looking for in response to that question.

Enterprise 2.0: Collective Intelligence

Jeffrey Walker and Stewart Mader of Atlassian spoke next on “Collective Intelligence: Monkeys or Memes?” (great title, making reference to the infinite monkey theorem), which was really about adoption patterns of enterprise wikis.

This is really going back to the theory that the IQ level of an appropriately organized collective can be greater than that of the smartest person in that group, and that’s the whole reason for using wikis in the first place, especially in a corporate environment, instead of just picking the smart guy to write the thing.

There’s some significant drivers for Enterprise 2.0 software, and they’re not all about the functionality: some are about the fact that it’s lightweight, easy to install (or software as a service, requiring no installation), easy to customize, and doesn’t require months for the IT department or a third-party system integrator to create a working solution. Many companies, however, still believe that anything that can be up and running in less than 6 months is just a toy; this attitude is driven by IT departments trying to hold onto their job security in a world where the new applications and tools cause an ever-increasing commoditization of their role.

Walker was a very engaging speaker, quite funny and lots of great material. He spoke about some of the advantages of enterprise blogging, whether purely internal or external-facing, and some interesting differences in how companies approach external-facing blogs: Sun just lets you go to town, whereas Cisco requires that you have VP approval and go through corporate communications in a process that must discourage many potential bloggers long before they’re (inevitably) turned down. He recommended checking out IBM’s blogging policy as a good balance for enterprises; having talked with a few IBM employees who also blog, I’ve heard the same thing from them.

Not for the first time this week, I’ve heard SAP’s developer network used as a great example of using blogs and wikis with their external community.

Pixar uses a wiki for project management of all film productions that they do; it started out in their IT and software development areas, but gradually moved into the business areas, which Walker feels is a typical adoption pattern. He also thinks that Enterprise 2.0 adoption is going to look a lot different in the next year than it has in the past year due to the ever-increasing momentum, market presence, and consumer awareness.

He finished up talking about Twitter, not just as a personal social networking tool, but as a platform that’s starting to be investigated by organizations like BART and LAFD to provide public service announcements via SMS. I’ve always seen Twitter as redundant with something like my Facebook updates or a my Skype status, but seeing some non-personal uses of it all of a sudden makes it really interesting.

Mader came up next to talk about some examples of what’s happened with collective intelligence. He’s the author of the book Using Wikis in Education, and used a wiki to collaborate with several others in order to move from material that he had published in his blog into full-on collaborative authoring. He also talked about how he boosted the level of collaboration by creating a Facebook group, which gained more members in a number of weeks than the number of readers his blog reached in several months; this really points out that Facebook is inherently a more social environment (duh) than the more passive activity of reading blogs, and the very act of someone adding themselves to the Facebook group would cause its presence to be extended to that person’s contacts, which is not true if someone is just reading your blog.

He took us through some of the content on wikipatterns.com, a site that Atlassian sponsors, which contains both people and adoption patterns and anti-patterns: another great resource if you’re considering an enterprise wiki and want to assist its adoption. He also talked about some of the challenges of enterprise wiki adoption: overcoming resistance to change, establishing the right scope, gaining trust amongst the contributors (usually manifested in questions such as “someone else can change what I wrote?” and “how can I approve edits?”), and embracing emergent behaviours and making them part of the corporate culture.

Mader addressed an issue straight on that I’ve seen with both blogs and wikis: the attitude that “if I put my expertise in a public forum, I’m no longer an expert”, or “someone will steal my ideas”. I’ve had this argument with several other independent consultants when trying to convince them to blog; it’s a little bit like an architect not wanting anyone to be able to walk through the houses that he designs in case they copy his ideas, when his real value is in both bringing those designs to life and developing new designs, not just selling the old set of plans over and over again. If the only thing that you will ever have to contribute is what you’ve already done in the past, then it’s time for you to retire.

There were some audience questions at the end about people’s need for attribution of material that they author; Mader feels that wiki editing history logs actually provide better attribution than an emailed Word document, and that the new generation of workers are more likely to be used to this form of collaboration. Attribution is an illusion anyway in this world of copy-and-paste; I’ve sent two documents to a client in the past several days, only to find that they copied the text out of my corporate template and put it into their own template before distribution within their company.