How Data Management Drives ‘Insights Engine’ Results
I just read a fascinating new article in the Harvard Business Review, “Building an Insights Engine.” It specifically focuses on Unilever’s reaction to global competitive forces among its many brands with an insights engine initiative that grew revenue at the $60 billion company by 4.1%.
The in-depth article offers a great perspective on how to go about systematically building organizational capability around the delivery of business insights—particularly about your customers (or patients, constituents, prospective voters, etc.).
The three authors, senior executives from Unilever and marketing agency Kantar Vermeer, open with two key insights about doing business in the digital age:
- Operational skills, such as superior manufacturing or distribution, are no longer an advantage. The new source of advantage is coming from customer centricity—insights derived from data.
- Scale is no longer an advantage. It’s not the amount of the data anymore, but the ability to connect the dots into actionable business insights faster than the competition.
“What increasingly separates the winners from the losers,” the authors write, “is the ability to transform data into insights about consumer’s motivations and to turn those insights into strategy.”
Why data management comes first
The article offers a ton of great information, particularly about using insights to shape strategy and drive successful execution. But I’m going to focus, as always, on the data management part of this discussion. First, I noticed that yet again, data management (part of what they call “data synthesis”) is the first of seven topic areas covered. In my opinion, data management needs to go first because the ability to integrate various streams of data is foundational to success.
How foundational? According to the article, research by Kantar Vermeer found that “67 percent of executives at overperforming firms (those that outpaced competitors in revenue growth) said that their company was skilled at linking disparate data sources, whereas only 34 percent of executives at underperformers made the same claim.”
Data connections are crucial to insight delivery. The best analytics tools, the most data, and the brightest people, will all fail to deliver business value unless you have an organization-wide data management strategy and architecture.
Integration isn’t easy
As fundamental as data synthesis is to leading companies, it’s not easy to pull off. Some of the key data challenges identified in the article include:
- Integrate massive and disparate data sets. You must be able to onboard any new data quickly, cleanse and enhance it (fill in missing data), and most importantly, find a way to relate it. To a customer, or a marketing program, or a sales or business initiative. Massive amounts of data are of little use if you can’t join the data sets to find new and interesting relationships. Good data integration and data quality tools will help here.
- Create one “version of the truth.” Say you want to gather all the information you have about a customer. You’re going to find that you have customer data scattered all across your systems, partners, channels—and, increasingly, social media. Collecting all of that information almost always results in duplicate, and often conflicting, data about the customer. The critical capability here is to be able to determine which systems are the most trusted and timely sources of data. At the end, you need to assemble an end-to-end picture of your customer that relates all of the best and most authoritative data available. This is a classic use case for master data management tools.
- Create (and manage) a common set of data definitions. This is so important, and so often gets overlooked. Data terms, definitions, and business context are critical to understanding and deriving insights from data. One of our customers ran into a classic “dueling spreadsheet” problem simply for lack of agreement around the meaning of different business terms. This gets even more difficult when an organization has multiple lines of business. Each BU thinks their definition is everyone’s definition. A shared and governed business glossary that is available to all who need it will help here.
These are critical data management capabilities that are foundational to the creation of a successful insights engine as described in the excellent HBR article.
One quick thing I would add to the authors’ perspective on data: Look for an integrated data management platform. You’ll need capabilities that will work with any data, across cloud, on-premise, hybrid and, increasingly, big data. Solving for one project or one department without thought of wider organizational needs will only perpetuate the “islands of data” problem that makes this kind of initiative so hard to deliver for many organizations.