Using Apache Spark? Learn more about the benefits of using Apache Spark on Qubole.
Learn More

A 5-Minute Guide on How NOT to do Big Data

October 22, 2015 by Updated June 16th, 2017

How Not to do Big Data

The verdict is in. Big data is delivering big benefits to businesses large and small. It’s little wonder that more and more organizations are anxious to dive into vast stores of data to extract hidden insights and gain competitive advantage.
But big data adoption comes with a caveat—do it right or don’t do it at all. And unfortunately, big data projects fail frequently because that warning is essentially ignored.

If your organization is eager to implement a big data analytics strategy, consider this 5-minute guide on how NOT to go about it—before jumping on the big data bandwagon.

In a nutshell, don’t do Big Data…

Unless management is 100 percent on board

A successful big data strategy requires total buy-in from management. Anything less than 100 percent commitment can dramatically reduce the effectiveness of the project, or derail it altogether. While you may think that everyone is onboard, statistics from Fortune Knowledge Group may indicate otherwise. In a survey conducted with 720 business leaders, over 60 percent indicated that when it comes to making complex decisions they tend to go with gut instincts, and trust real-world insights over hard analytics. If decision makers are unwilling to act on big data insights because they don’t trust the data, your big data initiative will fail.

Before going forward with a big data project, it’s imperative to make sure that all business leaders understand what the project entails and will give it their full support.

Without a business case

The theoretical benefits of big data analytics are compelling: personalized marketing, an enhanced customer experience, bigger profits, a competitive edge. However, when it comes to the practical application of big data, those hypothetical promises can become a little vague. That’s why it’s critical to build a business case before moving forward with a big data deployment. In creating a business case, it’s important to take a practical approach by looking past the potential benefits to the bottom-line and focusing on the real challenges, such as implementing an infrastructure to store and process massive volumes of unstructured data. After all, unless these fundamental processes are in place, companies won’t stand to benefit from the potentially game-changing insights that big data analytics can deliver.

If you don’t know what questions to ask

Getting caught up in the big data hype can cause companies to spend big and fast on big data before clearly identifying what business problems they hope to solve. There’s no question that the technical side of big data requires the kind of expertise in data science that business execs typically lack. But knowing what questions to ask of the data comes from having a deep understanding and working knowledge of the business. In order to extract real business value from big data, IT personnel and business leaders need to work together to make sure that the right questions get asked.

If you’re unwilling to break down silos

Many organizations compartmentalize their data, breaking it up and sequestering it in data silos, i.e., the manufacturing silo, the marketing silo, the sales silo, the accounting silo. But silos are bad news for big data. To yield optimal results, a big data initiative requires all collected data to flow smoothly across all channels of the organization. For this to happen, silos must be broken down to insure that all data is shared in a single unified hub where it can be freely accessed by all departments. Organizations looking to adopt big data should think twice if their policies discourage the breaking down of data silos.

If your organization can’t train existing personnel

Successful on-premise Hadoop adoption requires personnel with technical expertise to get things up and running, plus data expertise to make sure that the analytics side lives up to its full potential. Organizations that have the ability to nurture talent from within and train current personnel to make them equipped to handle big data challenges have a distinct advantage over companies that must hire the majority of talent from without. Turnkey big data services also help to remove this barrier by taking care of the setup and maintenance, so organizations only deal with the analytics side of the technology.

Given that today’s big data technology is primarily open source, and that a number of platforms are both affordable and readily accessible as cloud based services, organizations can take a trial-and-error approach to training their employees. In this environment employees are free to explore the data, starting small and learning as they go, until they acquire the necessary skills to do big data the right way.

Interested in learning more about the differences between big data platforms? Check out this complete vendor comparison.

  • Blog Subscription

    Get the latest updates on all things big data.
  • Recent Posts

  • Categories

  • Events

    Big Data World London

    Mar. 12, 2019 | London, UK

    Data Innovation Summit

    Mar. 14, 2019 | Stockholm, Sweden

    Spark + AI Summit

    Apr. 23, 2019 | San Francisco, CA

    Strata NY

    Sep. 23, 2019 | New York, NY

    Big Data World Asia

    Oct. 9, 2019 | Singapore