Talk with big data experts, and at least nine out of 10 will mention the value of democratizing big data use.
That’s because big data, properly utilized, is more than just a strategic initiative limited to a small section of a company. While using the wealth of data now at a company’s disposal certainly is a worthwhile initiative even in limited scope, it can serve a far greater cause if firms create a culture of big data use. The real power of big data starts to unfold when employees at all levels of a company leverage it and drawing insights from the data.
This does not happen overnight. Making big data accessible and more widely used requires several key actions on the part of management.
While there are many ways that businesses can encourage big data use, here are five of the most important.
1. Change the culture
Employees won’t reach for big data unless they know its power and how to harness it. This comes from training, and making a point within the company to promote big data use for company questions of all types.
“Build a culture of big data analytics,” stresses Theo Vassilakis, who spent nearly eight years handling Google’s big data needs before founding his current firm, Metanautix. “Educate your people about the data and encourage empiricism, scientific thinking, and iteration. There are many ways to do this including competitions, hackathons, industry collaborations, management participation, awards/rewards, etc.”
Creating the right culture for widespread big data use also means promoting a bias to action. When employees need data, they need to know how to go out and get it instead of waiting for the data lake to reach them.
2. Make self-service the focus
Key to democratizing big data use is enabling a self-service model where business managers have both the data and the fit-for-purpose tools needed to answer questions, according to Rishi Kumar, chief scientist for analytics firm, eBench, and the former director of analytics for Unilever.
“Shift the role of the expert analyst from solving problems directly on an ad hoc basis to developing self-service analysis consisting of data, metrics and data visualizations that enable general analysts and business managers to solve problems on an ongoing basis,” he says. “By activating your data program using your entire workforce, you can achieve full leverage of your data and specialist resources and reduce your dependence on high-cost labor.”
3. Launch a big data help desk
Even with training, most employees won’t likely become data wizards. And that’s okay. More important is empowering workers to look for answers from big data, and to ask for help when the tools at their disposal are not enough.
“Democratizing data is about creating an environment where people can openly ask business questions and analysts with skills are constantly in contact with the data they need to answer those questions,” notes Kumar. “As a result, a business obtains a sort of ‘wisdom of crowds’ and leverages its data and analyst resources to its full potential.”
To that end, setting up a big data help desk can be a smart move. Just as employees know to contact the IT department when they have a computer question or problem, a data help desk can serve as a complement to employee self-service data analytics options.
4. Enable visual self-exploration of data
Big data analytics is about seeing relationships and outliers in large quantities of data, and most employees are visual creatures. So one way to boost data use is making it easy to visualize this data.
A big part of making self-exploration of data visual is making it intuitive and easy. Data visualization tools must be simple and responsive, without a vicious learning curve. Employees need to be able to work with the data at the speed of thought, so overly complex visualization tools just don’t cut it.
Make sure that there are easy options for sharing this data, too, so employees can email findings to colleagues, include it in PowerPoint presentations, and output it in a number of formats for further use.
5. Assign different data to different groups
Enabling employee data use and exploration requires data access, of course, but access should not be conceptualized as one-size-fits-all.
A data scientist might be given complete access to raw data, as well as advanced tools for manipulating the data. A manager, on the other hand, might benefit more from a curated set of data and tools that are simpler if also less powerful.
“Contrary to popular belief, democratizing data is not about providing everyone in an organization access to all data,” says Kumar.
Making data more usable by employees does not imply a free-for-all. Data housekeeping matters more than ever when a business opens up its data stores.
There is the security and privacy issue, too, of course.
“It’s not quite the gold rush anymore,” says Vassilakis at Metanautix. “The cost of making regulatory compliance, security and privacy mistakes is higher than it’s ever been. Build governance, audit and security into your big data projects from the start.”
Do open up your data, though. Better data use is becoming a key competitive battlefield, and that won’t change any time soon. Opening up data access, and getting employees to leverage it more, is central to making the most of a company’s advantage.
Author: MaeMae Kowalke