Over the last decade, the collection and use of data by companies, governments, and other entities to inform every critical decision they make has grown by leaps and bounds. Along the way, new technologies and techniques have coalesced into the field we now refer to as big data, creating whole new ways for all that data to be put to use. That has helped fuel a wave of investment by big businesses into big data technologies – mostly due to fear that failing to do so will leave them at a disadvantage compared to their rivals.
The results haven't been pretty. Some estimates put the failure rate of big data projects as high as 85%, and surveys continue to report that companies still face an uphill struggle in their efforts to become data-driven. Those results indicate a massive disconnect between business spending and the hard realities of the big data revolution.
The bottom line is that it takes an organization-wide effort to make big data initiatives work, and that requires buy-in at all levels and some specific steps to prepare a workforce to execute the necessary strategy. As it turns out, it may not even be as difficult as some companies are making it seem. Here's what businesses need to do to turn their organizations into lean, data-driven machines.
Get Organized
The first step to creating a data-driven organization is to get a handle on what data the company already has access to, where it's stored, and what it's currently used for. The big hurdle that most organizations face in this task is that the data they have will likely be stored in different systems that can't interact with one another. These data silos make it difficult to assess and categorize a company's data, making such an inventory a long and arduous process. That's one of the reasons that many businesses fail to accomplish this foundational task – it's long, boring, and difficult. To succeed in becoming a data-driven organization, however, it's essential to conduct a complete and thorough data inventory that makes clear the scope of the business's data assets.
Choose a Platform
The success or failure of any big data initiative will rely on the data you've got to work with, but also the tools that are chosen to do the job. In creating a data-driven organization, this is not a choice to be taken lightly. It's important to make strategic investments (which can be significant) into a data science platform that will become the bedrock of all future data-driven efforts. That means it's essential to choose one that meets not only the data needs of the organization today but also it's anticipated needs for the future. Although individual business needs may vary, choosing the right data science platform generally comes down to programming language compatibility, availability of visualization tools, and extensibility. Getting the choice right can go a long way toward bringing data into every aspect of the business while getting it wrong can create a roadblock to data use at all levels of the business.
Spread Data Literacy Throughout the Organization
Believe it or not, building the structure to generate data-driven insights for a business is the easy part of creating a data-driven organization. The hard part is getting employees at all levels to understand and accept the use of data in their everyday roles. To make that happen, it's a good idea to start by naming a Chief Data Officer (CDO) for the business who will oversee the data literacy training efforts that must happen throughout the organization. If selecting a candidate from existing management, it's a good idea to have them complete an online master of data science to give them the high-level knowledge they'll need to do the job right.
For everyone else, creating a mandatory data literacy training program is an excellent way to impart the necessary data skills into the workforce. For a model of how to do so, companies can look to vacation rental giant Airbnb, who has built just such a program and is seeing improved data adoption throughout the company among those who have already completed the training. This is quite possibly the most important step in the process of becoming a data-driven organization, because it empowers individual employees to use data every day, and normalizes it in a way that makes it second nature to them.
Data at Every Level
With the right data, tools, and training in place, any business should be able to transform itself into a data-driven organization and escape the dreadful results that have come with so many early efforts by others. That's not to say the process won't be painful. To begin with, getting business data centralized, cleaned up, and ready for use can take months, if not years. The same goes for the training efforts that must accompany any data initiative. Results almost never happen overnight, so the most important thing that companies can do is to have faith in their data program and see it through to the end. The result should be an organization that knows how to use data at every level and does so to support a thriving, true 21st-century data-driven enterprise.
