Two back-to-back morning keynote sessions at the first day of the Gilbane Conference in Boston featured a wide variety of presenters speaking on an array of information and content management topics. Following is a roundup of the key points each keynote speaker made.
Brent VerWyst, Product Manager, Enterprise Search, Google
While enterprise search market observers differ on whether enterprise search is undergoing commoditization or disruption, VerWyst made a solid argument on the side of disruption. He explained how enterprise search is using algorithms to understand language enough to make relevant search suggestions. For example, a Google search for the term “diffraction grates” will use word stemming to come up with the likely suggestion “diffraction grating,” a term used in mass spectrometry.
“These are the types of things we spend a lot of time on to expand the adoption and use of search,” said VerWyst. He also said that cloud computing is disrupting search in a “back to the future” way. “Traditional content in file shares and applications is migrating to the cloud,” he said. “Documents link to other documents in a rich structure that starts to look a lot like web pages.”
MaryLee Kennedy, Senior Associate Provost, Harvard University
Kennedy said organizations should create “regenerative environments” that nurture ideas which lead to other ideas. As an example, she said most schools at Harvard allow students conducting research to hyperlink their notes to a global, open source database that is then made available to the Harvard library search service. Users can also place notes and research findings from the database onto other social platforms such as YouTube and Twitter.
“You create metadata from (documents in the database),” said Kennedy. To date, Harvard has created more than 12 million pieces of metadata from this system which are available through open source and have been used more than 1,000 times.
Jonny Kaldor, Founder, Pugpig
Kaldor, creator of the Pugpig HTML publishing platform, said the “paperless world” which pundits have been incorrectly predicting for years is finally within sight thanks to the development and widespread adoption of tablet computers. “We finally have a viable alternative to paper,” he said. “A tablet is heavier than a piece of paper but a damn sight lighter than a library.”
Kaldor said the New York public schools are experimenting with a program that puts all learning materials students use throughout the day on tablet computers. In response to the growing importance of the tablet platform, he said publishers must make content the “heart” of what they do. “Content ends up as bitmaps on a page and you can’t leverage or syndicate it,” he commented. “Designers must become developers who can use content management systems with forms and fields.”
Dr. Bill Simmons, CTO/Cofounder, DataXu
Simmons, who helped found the DataXu digital media management platform, said Amazon Web Services and the development of open source tools like Hadoop have made solving the Big Data challenge cheap. “Two guys in a garage can have access to 1,000 servers and process Big Data in a couple of days,” he said. “You don’t have to work at IBM or Google to innovate with Big Data.”
In this environment, DataXu allows users to optimize and automate media buying using end-to-end Big Data analytics, helping marketers satisfy CFOs who likely only care about ROI and not branding or other “soft” returns. As an example, he said after marketing its compact, inexpensive Fit model to drivers younger than 25, Honda discovered it was also appealing to senior citizens. “This was a missed opportunity” marketers could have capitalized on with automated Big Data analytics, according to Simmons.
Gerry Murray, Research Manager, CMO Advisory, IDC
Murray suggested that consumers should “own” their behavioral and demographic data and willingly supply it to marketers in exchange for content filtering tools that would allow them to only receive promotions they are interested in. “There would be an ecosystem of outside influencers to help marketers deliver content,” said Murray. “Content differentiators would be context and source.”
Consumers would be less aggravated by unwanted advertisements while marketers and advertising platforms would drive greater ROI. In one example, Murray said this type of individual targeting would allow a TV network to sell a single Super Bowl ad slot thousands of times to advertisers targeting different groups.
Stephen Powers, VP/Research Director, Forrester Research
Web content management (WCM) systems and more established enterprise content management (ECM) systems actually have much in common, according to Powers. Both present the “utopian dream” of having a single content repository with a single set of solutions and services from a single vendor sitting on top, but in reality do not and never will deliver upon this potential.
“WCM is an integration play,” said Powers. “Nobody has an end-to-end customer experience management system. Some vendors have a lot of the components, but none have all.”
Tony Byrne, Founder, Real Story Group
Byrne said mobile and cloud technologies offer the “notion of freedom and liberty,” with customers and employees interacting with the organization from anywhere they want. In addition, he said the business thinks mobile and cloud will free it from having to work with “those guys in IT with the black fingernail polish.”
In reality, Byrne said properly utilizing enterprise cloud and mobile technology requires the business and IT sides to work more closely together than ever, but in a different manner. “You can’t say, ‘I’m finished with IT, I’m going cloud,’” said Byrne. “Align business with IT. Work through the cloud, rather than escape to the cloud.”
Scott Liewehr, President/Principal Analyst, Digital Clarity Group
Liewehr presented a contrarian view that while Big Data does provide value in some situations, there is no “mandate” to use it. “There is a snowball effect where you create a solution without a problem to solve. A few years ago everyone was building portals, but there were plenty of failed portals that didn’t solve true business problems.”
In many cases, Lieweher said small samplings of data give all the accuracy you need. “Most political polls from the recent elections were pretty accurate using sample groups of 450 people,” he said. “A lot of decisions are based on simple data -- do you need to collect the universe?”