7 Ways to Help the Business Digest Big DataFebruary 13, 2013 1 Comment
By Joe McKendrick, Author and Independent Researcher
In a recent tweet, Kirk Borne, a data scientist at George Mason University, pointed out that “only four words in English end in ‘dous’: tremendous, horrendous, stupendous, hazardous. And all describe Big Data hype cycle!”
Big Data is now the term du jour, thrown about at conferences with wild abandon, often sneered at, but more often that not greeted with a mix of nervousness and fascination. The term many be overused and amorphous, but it’s something everyone is now dealing with at one level or another. Indeed, Big Data is “tremendous, horrendous, stupendous, hazardous,” all at once. If it isn’t the volume – such as terabytes’ and terabytes’ worth of data – then its the variety enterprises are seeing, especially with unstructured files such as Word documents, PDFs, social media and weblog data.
But all the data in the world has no value unless it’s reaching, in an actionable form, decision makers in the organization. With that in mind, here are some nuts-and-bolts ways to give Big Data its heft:
1. Unify data access: Big Data tends to be “inaccessible to the organization as a whole because it only exists in the domain in which it was created or stored, and often it is difficult to access even within that domain,” says a report out of TSIA and Coveo. Organizations need to figure out ways to break down the silos and provide a single view of all relevant data. One way to do it through a unified service layer, structured similar to a service oriented architecture. Search technologies also are an enabler of unified enterprise views, the TSIA/Coveo paper points out.
2. Separate “information” from “data.” That’s the word from Dr. Michael Wu, principal scientist of analytics at Lithium, who says that too many people think the data itself is a valuable commodity. However, it is not the same as “information.” He says the qualities that separate the gold nuggets of “information” from the mountain of data include interpretability, relevance and novelty. “Information must be novel to be insightful,” he says. “That means it must provide some new knowledge that you don’t already have.”
3. Encourage analytical skills – for everyone. Don’t confine analytics to a group of quants tucked away in some corner of the organization. A recent survey I conducted for Tableau Software and Unisphere Media finds only 10% of organizations offer self-service BI on a widespread basis. Individuals capable of looking at the data in new and innovative ways may come from anywhere, and should be encouraged. In terms of presentation, information needs to “pop”—not get buried in rows of numbers or figures as they have traditionally been in bland reporting tools and formats. BI dashboards, portals and other interfaces need to include graphs, dials, and other representations of data covered.
4. Enable self-service analytics against Big Data. To really promote analytical thinking at all levels, deliver insights on a self-service basis. Business users – who are used to the speed of Google queries on the web – are very frustrated with the glacial pace with setting up internal corporate business intelligence with their IT departments, then only getting limited reports in PDF formats. What is needed is the ability to do rapid and flexible data querying across multiple data sources.
5. Make Big Data insights available via apps. Once self-service BI is a reality, why can’t decision makers view important details on their smartphones? I once heard Competing on Analytics guru and best-selling author Tom Davenport wonder out loud at a conference why there weren’t more analytics being made available as a “cute little app.” By offering analytics through simple, single-purpose mobile apps, decision-making can be brought into a whole new realm. A report issued last year by Dresner Advisory Services says some functions are already starting to be made available through simple little apps, such as key performance indicator marketing, drill-down navigation, data selection and filtering, and alerts.
6. Manage Big Data cost-effectively. It’s not worth it to try to extract, transform and load data through a big honking data warehousing type of environment – the process is just too costly. Instead, look to lightweight, open source tools that will compress Big Data stores into management file systems, such as Apache Hadoop. Hadoop also has an impressive ecosystem of tools – or subprojects – including Sqoop (SQL-to-Hadoop), a data import tool; Hive, a Hadoop-centric data warehouse infrastructure; and Pig, an analysis platform for data from parallelized applications. Those are just three.
7. Get business users involved, and push the boundaries. Big Data opens up opportunities to ask questions that were never even possible to consider within standard, relational data environments. Many BI tools offer solid user interfaces, but business users need to be educated on how to make sense of large volumes of data, how to manipulate the data to answer business scenario questions, and how to create documents to communicate insights in a meaningful way. Once they do, they produce amazing new forms of value. Remember, today’s successful organizations are no longer just manufacturers, technology providers, or government agencies – they are all now data companies.
Joe McKendrick is an author and independent researcher, covering innovation, information technology trends and markets. Much of his research work is in conjunction with Unisphere Research/ Information Today, Inc. for user groups including SHARE, Oracle Applications Users Group, Independent Oracle Users Group and International DB2 Users Group. He is also research analyst with GigaOM Pro Research.
He is a regular contributor to Forbes.com, and well as a contributor to CBS interactive, authoring the ZDNet “Service Oriented” site, and CBS interactive’s SmartPlanet site.
Joe is a co-author of the SOA Manifesto, which outlines the values and guiding principles of service orientation in business and IT.
In a previous life, he served as communications and research manager of the Administrative Management Society (AMS), an international professional association dedicated to advancing knowledge within the IT and business management fields. He is a graduate of Temple University.Analyst Blog, DATA and ANALYTICS