Data Informed

Business executives want their companies to quickly adopt new big data technologies, but CIOs know implementing these projects is complex. says Martha Heller, author of a new book about IT leadership. In this podcast, Heller discusses how CIOs need to manage what she calls the paradoxes of their role and how this effects their work on big data projects.

With the rise of predictive analytics, statistical computing software like SAS, R and SPSS are playing a vital role in driving fast analysis of large data sets with complex algorithms. Of those three statistical languages, R is the only one that is open source. SAS is the largest privately held software company in the world, and SPSS is now shepherded by IBM. Revolution Analytics provides an enterprise distribution of R. The company is looking to battle both SAS ans SPSS on price, performance on distributed and massively parallel computing, and adoption in the cloud, according to Davis Smith, Revolution's vice president of marketing and community. Smith admitted it is an uphill battle. SAS, a company that was founded 40 years ago, is used by most Fortune 500 companies. But Smith said because R is open source and costs nothing to download, it's widely used in academic settings and most new graduates with statistics and analytics degrees have used R. Revolution Analytics has also focused on connecting R with other companies, including IBM and Jaspersoft, to provide business intelligence platforms, connect to Hadoop, NoSQL databases and data in the cloud. The resulting ecosystem, Smith said, is as robust as any all-in-one data stack, he said. In this interview with Data Informed Staff Writer Ian B. Murphy, Smith discusses how R was started and why it continues as one of the longest open source projects, the additional features Revolution Analytics' distribution has compared to stock R, and how it's battling with SAS to become the most used statistical computing language. (Podcast running time: 21:41.) Email Staff Writer Ian B. Murphy at ian.murphy@wispubs.com. Follow him on Twitter @IBMurphyatDI. Check out other podcasts from Data Informed. The podcasts are also available on iTunes.  See related stories on Data Informed: • 4 Tips for Writing an Analytics Job Description • North Carolina State's Institute for Advanced Analytics Trains Analysts for Jobs in Business • SAS CEO Jim Goodnight on Modernizing Analytics for Massively Parallel Systems

In this podcast, MapR Technologies CEO John Schroeder said creating a framework to use SQL is the next step for Hadoop because the ability to have interactive data queries is essential. He also discusses which jobs are in the greatest demand, and what knowledge is crucial to work with Hadoop.

  A beautiful piece of art based on data can be a joy to look at, and be food for thought. But the boss doesn't want food for thought, he wants a signpost pointing the way forward. Lucas Walker, a co-founder of Venngage, said business users need to create visually appealing reports that provide suggestions for the next best action. If they can make those data-based visualizations without having to involve IT, all the better, he said. That's the idea behind Walker's company, which spun out of the popular visualize.me resume infographic project. In this interview with Data Informed Staff Writer Ian B. Murphy, Walker discusses what makes a good data visualization, the difference between an infographic and report, and why potential customers of a data visualization platform like Venngage are clamoring for real time and predictive analytics. (Podcast running time: 14:10) Email Staff Writer Ian B. Murphy at ian.murphy@wispubs.com. Follow him on Twitter @IBMurphyatDI. Check out other podcasts from Data Informed. The podcasts are also available on iTunes. For more on data visualizations, see: • Facebook Adopts Tableau for Analytics Dashboards • Relay Technology Combines Analytics, Search and Visualizations to Bolster Drug Discovery • Telling Stories with Visualizations: Lessons from Data Journalists

Two of the best ways to ensure Hadoop's wide adoption in enterprise: Partner with data industry giants that have products most companies are already using, and make the distributed data storage and computing system 100 percent open source. Those are the strategies that Herb Cunitz, the president of Hortonworks, is pursuing to make sure Hadoop succeeds in the top businesses in the world. When Hadoop succeeds, Hortonworks does too, he says. Cunitz said that making Hadoop truly open source, where every improvement is available to the community, will prevent users feeling locked in. It avoids players in the Hadoop ecosystem from having to create complicated connectors, or products that don't work with other distributions. By pursuing this approach, Cunitz says his company can focus on service and support, and on creating more innovative tools to enrich Hadoop. In this interview with Data Informed Staff Writer Ian B. Murphy, Cunitz discusses Hadoop's current maturity level, what the landscape will look like for Hadoop adoption in 2013, and why Hadoop should herald a new era of 100 percent open source development. (Podcast running time: 17:40) Email Staff Writer Ian B. Murphy at ian.murphy@wispubs.com. Follow him on Twitter @IBMurphyatDI. Check out other podcasts from Data Informed. The podcasts are also available on iTunes.  Find more articles and podcasts on Data Informed about Hadoop.  

Next year many major companies will move their big data projects from the planning stages to implementation, with new industries and lines of business inside enterprises getting in on the action. That's just one of several predictions for big data in 2013 from Nancy Kopp, IBM's senior program director of product, marketing and strategy for big data. Kopp said that in order for IT to keep up with new big data initiatives, the new focus will have to be on agility and time to insight, surpassing even performance as the top necessity in data warehousing. The days of the "monolithic data warehouse" are gone, she said. In this interview with Data Informed Staff Writer Ian B. Murphy, Kopp discusses what next year's data scientist will be required to do, how  technologies like Hadoop will grow and mature and how investments in IT will shift as enterprise goes from searching for value in their data to expecting fast insight. (Podcast running time: 19:19) Email Staff Writer Ian B. Murphy at ian.murphy@wispubs.com. Follow him on Twitter @IBMurphyatDI. Check out other podcasts from Data Informed. The podcasts are also available on iTunes. 

Omer Artun left Brown University with a Ph.D. in computational neuroscience and physics with an excitement for predictive analytics and data mining. After consulting at McKinsey & Company, he developed a deep appreciation for data-driven decisions. But after stints applying marketing analytics at CDW/Micro Warehouse and Best Buy (where he was senior director of B2B marketing), Artun came to realize there weren’t good tools to help marketers and salespeople tame the massive volume and variety of data. He wanted something to recommend actions based on that wealth of information that would help his companies start a relevant relationship with customers. Related Stories Check out other episodes in Data Informed’s podcast series.        Read more» “I saw firsthand that using data and analytics, you can use much better decisions,” Artun said. “But I saw that spreadsheet-level analytics was not enough when you have a lot of data.” In 2006, Artun began building AgilOne, a cloud-based marketing analytics service that manages data and uses predictive analytics to make clear-cut marketing campaign recognitions. On Nov. 28, AgilOne launched its core product for general release and announced it had received $10 million in additional venture capital funding. Please listen to the excerpt of a conversation between Data Informed staff and AgileOne CEO Omer Artun, where he discusses the importance of data management in marketing, how predictive analytics can help sort through all the noise to find insight, and why it’s important to create tools that recommend concrete actions. Email Staff Writer Ian B. Murphy at ian.murphy@wispubs.com. Follow him on Twitter @IBMurphyatDI.

Python's growing programmer community believes the simple coding language is the next big open-source project to work with big data. Peter Wang of Continuum Analytics talks about supporting the open source development of Python in this podcast.

Organizations who are hiring data scientists should avoid a few common mistakes like requiring Ph.Ds, focusing solely on skills or hiring arrogant data rock stars. Talent Analytics CEO Greta Roberts said curious minded people are often self taught, and don't get advanced degrees. Communication is an important skill, but not every data scientist needs to be a great communicator. Business sense, Roberts said, is much more important. These are some of the insights gleaned from the survey of data scientists and thought leaders Talent Analytics conducted over the summer. The survey, called the 2012 Analytics Professionals Study, will be available in full in December and will help businesses understand and value the essential data science skills and traits in their employees and recruits. In part two of this interview with Data Informed Staff Writer Ian B. Murphy, Roberts discusses what elements of business acumen are important for a data scientist, what to look for when interviewing for a data analyst position and what traits aren't important as they seem. (Podcast running time: 13:35) For related articles from our archives, see the following: Rethink your organizational chart for analytics teams. Building an analytical culture for big data. Assembling a big data team.

The global analytics talent shortage makes creating a well-rounded data science department a daunting task. Personalities collide, skill sets vary and a traditional interview is not always an accurate determination of whether a professional will fit the position. Talent Analytics CEO Greta Roberts said that’s why she turned an analytical lens back on data scientists to measure important characteristics found in the successful professionals in the field. Roberts said the 2012 Analytics Professionals Study, which will be available in full in December, will help businesses understand and value the essential data science skills and traits in their employees and recruits. In part one of this interview with Data Informed Staff Writer Ian B. Murphy, Roberts discusses some of the pieces of the study: how to build an analytics team, what characteristics to look for in prospect, and how to make connections with leading analytics university programs. (Podcast running time: 13:59) For related articles from our archives, see the following: Rethink your organizational chart for analytics teams. Building an analytical culture for big data. Assembling a big data team.  

Companies looking to extract value from their data need to first think about creating an environment where business data is clean and workable. Dirty data leads to questionable results. But IT departments will hit major obstacles if they try to install a master data management (MDM) program without first working with business users to figure out goals and requirements of data governance projects, according to Dennis Moore, senior vice president and general manager of master data management at Informatica. Like any other IT project, it's crucial to be able to show off a few victories before taking larger MDM projects up to the executive level, Moore said. Trying to "boil the ocean" is to doom a new, ambitious project to failure. In this interview with Data Informed Staff Writer Ian B. Murphy, Moore discusses why MDM is key for creating insight from enterprise data, how to get a MDM project off the ground and what obstacles to avoid to ensure success. (Podcast running time: 17:09.)

Analytics aren't just for analysts anymore now that businesses are buying in to the value of their data. So SAS, a company that has made its mark in statistical analysis software, is connecting its products to a visual user interface designed to allow non-technical business users to work with data, too. Related Stories SAS emphasizes speed in new analytics server, talks up visualization interface.        Read more» SAS vice president of sales development and project management Randy Guard showed off an improved user interface and more capabilities for the Visual Analytics application last week at the company's Analytics 2012 and Professional Business Leadership conferences in Las Vegas. Guard said the new UI is meant to be "walk up and use" easy on the level of popular Web apps like Google and Facebook, because that's what customers have come to expect. Guard said coupled with the company's High Performance Analytic server, which creates a massively parallel processing environment that allows the Visual Analytic software to run SAS's complicated statistical models, the new UI will be the graphcial face of the company for several years to come. In this interview with Data Informed Staff Writer Ian B. Murphy, Guard discusses the Visual Analytics product, the announcement that the High Performance Analytic server will run on Hadoop, and why it's important to create visualization software with mobile devices in mind. (Podcast running time: 14:38)

Modern analytics are fast, no matter the size of the data you're crunching, so analysts can run several iterations of their models and reach insights more quickly than using systems of the past. In the past three years, Jim Goodnight, the CEO and co-founder of SAS, has been pushing his company to modernize the software algorithms it began developing 35 years ago. The goal: to create a massively parallel processing platform to run analytical jobs faster than ever before. Related Stories SAS emphasizes speed in new analytics server, talks up visualization interface.        Read more» The result is SAS's High Performance Analytic server, which is capable of running complex models on billions of rows of data in seconds. The company announced at its Analytics 2012 conference in Las Vegas on Oct. 10 that it would roll out optimized versions of its most popular software, ready to run on an analytical appliance built on Greenplum, Teradata or Hadoop, every six months. In this interview with Data Informed Staff Writer Ian B. Murphy, Goodnight discusses the complexities of rewriting statistical algorithms for massively parallel processing, the boost in productivity that fast analytics provides, and the benefits of running models on a whole data set instead of a data sample. Goodnight also discusses the thinking behind SAS's new Visual Analytics and how it ties in to the in-memory appliance. (Podcast running time: 17:26.)  

To the old saying, that there are three kinds of lies--lies, damn lies and statistics--Richard Dale asserts there is a 21st century version. Dale would say that there are lies, damn lies and visualizations. Dale, who used to work for Sigma and Parnters, is now managing director of Big Data Boston Ventures. He recently moderated a panel of visualization expertsfor the Massachusetts Technology Leadership Council, said anyone looking to persuade someone to one side of an argument can throw together a data visualization. But for a visualization to effectively and accurately reflect a truth, it requires hard work, teamwork and a rigorous iterative process. In this interview with Data Informed Staff Writer Ian B. Murphy, Dale discusses why he feels business culture needs to change to better understand data processing and analytical methods, what elements make up a powerful data visualization, and why today's tools for analytics and visualizations are only the beginning of a potential paradigm shift. (Podcast running time: 16:23.)  

Most companies using smart phones and tablets are just now realizing how powerful they can be when you develop applications tailored to their needs. Top level executives are interacting with data in ways they never have before, leading to better informed conversations with their BI teams, and developers are creating applications that take advantage of the wide array of sensors that come standard in most leading mobile devices. MicroStrategy CTO Jeff Bedell and Dan Kerzner, the company’s senior vice president for mobile, said mobile analytics have "reached the tipping point" where these devices are so powerful that the biggest limit on new use cases is the imagination of the developers. Whether it's just moving a business process forward by allowing manager approval from anywhere, or using the camera, location services and the accelerometer to create new data sets, mobile devices are changing how business is done in the field. In this interview with Data Informed Staff Writer Ian B. Murphy, Bedell and Kerzner discuss how tablets and smart phones are used to enhance meetings today, and what some uses that are likely to emerge in the near future, and why they are so bullish on mobile analytics development over the next 12 to 18 months. This is the second half of a two-part conversation. (Podcast running time: 15:46.)