Skip to main content

How retailers overcome the 'silo problem' in data management

Paul Newman Archive

The retail industry is one area in particular that's taken a keen interest in getting more out of customer data. If businesses have more information that's useable for big data analytics, they can make improvements in everything from their pricing plans to their in-store merchandising strategies. There's seemingly no limit to their capabilities.

Unfortunately, retail is also an area where it's practically impossible to house all of a company's data under one roof. Here's the problem: People buy products using a wide variety of channels. They're taking advantage of everything from brick-and-mortar store locations, to the phone and postal mail, to website and mobile shopping apps. All of these avenues have different ways of doing business - and as such, they collect different data, too.

This presents numerous problems. For one, it's difficult for companies to take the data they're collecting, put it all in the right place and keep it on hand for future use. It's also very complicated to discuss data quality, as each type of customer information is gathered in a different way and comes tagged with its own attributes.

This is what's known as the "silo problem," and in retail, it's especially worrisome. Merchants want to do everything they can to overcome this difficulty, but there are unfortunately a lot of logistical hurdles that come their way.

Breaking down the silos is a hassle, according to The Multichannel Merchant. But Sandra Gudat, president and CEO of marketing agency Customer Communications Group, told the news source that fortunately, there are things that companies can do to tackle this problem.

"When your data is siloed, it can be an overwhelming task to ensure updates are made quickly and consistently across the board," Gudat stated. "That's why the best practice is to have a 'single source of truth.' You need a master customer warehouse that is the central hub for customer data and prioritizes the information that flows back and forth between the various sources."

Below are the four general steps that companies take to move past their organizational issues:

Make an interdisciplinary effort
A common problem in data management is that companies have their information management personnel scattered across multiple departments. Their IT, human resources, marketing and sales people all have data in their own separate places, and no one's able to work together. One key goal should be to have companies work on data in an interdisciplinary way, sharing objectives and responsibilities across the whole office.

Put the right people in charge
Of course, while it's important to have many hands on deck, it's also good to have one person (or perhaps a group of people, in a larger enterprise) in charge, coordinating the overall effort. Gudat uses the term "data steward" to describe someone who understands data management and oversees the whole warehouse. Whatever you call it, make sure you have someone capable of running the whole show.

Devise a business-oriented plan
Overcoming the data silo issue is a complicated problem, but it should begin by identifying your overall business objectives and assessing how data helps. What are your company's broad goals? Do you want smarter marketing plans, better customer service or simply more profits? In any event, the best plans begin with the company's collective mission in mind.

Map out the right "data flows"
Your plan should be one that's well organized and utilizes the strength of all your best people. Make sure that you've clearly mapped out all data management processes ahead of time - you want to know where information is supposed to "flow" as it's collected and disseminated among the corporate ranks. Planning and organization can help any company overcome its troubles with data quality and management.