Big Data is cool. The Internet of Things is cool. Machine Learning and Artificial Intelligence are cool. Predictive and Prescriptive Analytics are cool. Having bad Data Quality so that none of those cool technologies actually work is not cool. A coherent Data Strategy that involves a comprehensive Metadata Management plan, well-defined Data Governance and Data Stewardship, and the ability to leverage all an enterprise’s disparate data assets is something many (if not most) organizations only dream about. “How do I really use data to transform my business? It’s a question I get all the time from clients,” said Donna Burbank, the Managing Director at Global Data Strategy, in a recent DATAVERSITY® interview. “With data today driving so much business transformation, it’s become necessary to look at the business first and seeing how the business can drive the technology rather than the other way around.”
Often when technology people, or data people, go into a business meeting and start discussing terms like Data Governance, or Data Strategy, or Metadata, or MDM, or any number of other traditionally “techy” terms, faces go blank and people start checking their phones. If those same data people start discussing business definitions, or start the conversation around definitions of “customer”, or how Sales calculates total sales by region, faces expectantly light back up. “To them, that’s what it’s about,” she said. “It’s about context. What does it mean when I say ‘customer’, or ‘product’, or ‘account’?” Data Strategy needs to begin with Business Strategy, because once a business sees the value of high quality data (and they already do if they are asked in terms they understand) then the rest is a matter of communication and implementation.
Transforming the Business
Today’s world is full of new data. Companies can now collect social media data, open data, sensor and geo-spatial data, and more. They can collect surprisingly specific data, such as foot fall data, which is data collected from smart phones that shows travel patterns of people as they move around a town or within a store:
“But what it comes down to,” said Donna Burbank, “is how can I use that data to optimize my business processes? How can I do what I am already doing better? How can I optimize my marketing campaign by getting a 360 degree view of ‘customer’? Not only can I enter the internal data to try and get that cohesive view, but also external data.”
Organizations today have the ability to get a true 360 degree view of their market segments, customers, and product movements. Businesses can transform the way they understand themselves and their place within the market. Energy companies can use smart meters, connected through the Internet of Things, to better understand the energy usage patterns of all their customers – while at the same time a single customer can log in to their smart home via the Internet of Things and shut off their heat while they are at work or on vacation. Brick-and-mortar stores can track the way customers move around their displays and improve their product placements based on those traffic patterns.
None of those examples can work without a comprehensive Business Strategy that aligns itself within a cohesive Data Strategy. “This has all changed from 20 years ago when I first got started in Data Management,” said Burbank. “The business now understands that data is there to help transform the business. The call me asking about how to align both together; they know they can drive innovation through data, but they just don’t know exactly how.”
The starting point is Business Strategy. “We talk business before we talk technology,” she said. “Our clients come to us and say they know that they are stuck. I had one client say they heard that Big Data is cool, but they had no idea how to make sense of all of the information about it.” The focus needs to not be on the Big Data, or the implementation of MDM, or the fancy new BI platform, or any of the technology side – those elements can fall into place once the big questions are answered: What does the business need? What are the business’ pain points? What can data (and thus technology) do to mitigate those problems and provide the means for Data Strategy success through business-driven initiatives?
Focus on the Business Need
In a recently published article titled “The Business of Data: Focus on the Data that is Important to Your Business,” Donna Burbank wrote that:
“A core aspect of any successful Data Strategy is to prioritize the data that is most critical to the business, and focus comprehensive efforts around this data first. While this seems obvious, it’s often overlooked. While I’m a huge supporter of information management frameworks and maturity assessments, a downside of these frameworks is that, by definition, they have a strong focus on the technical components necessary to implement a comprehensive data strategy such as Metadata Management, Data Quality, etc. As a result, many organizations use them as a ‘checklist’ to prioritize their efforts.”
She is not saying that such frameworks and technology implementations are not important pieces of the entire strategy, but rather that the initial focus needs to be on why those pieces are important. In her interview, Burbank gave the example of a client discussing the fact that they had 16 different versions of how they calculated total sales and 20 versions of “customer”. “The business already understands this,” she said. “They may not use terminology like Metadata, or Data Governance, or Data Quality, but they know that they have bad data and that it is adversely affecting their day-to-day business operations.” Their technology may actually be working very well; their Metadata Repository could be an amazing implementation, along with their ETL processes, and probably even their new Hadoop integration – but without a clear understanding of the business issue all of those technologies are inoperable:
“The end case might be MDM, or a new Data Warehouse, or Data Quality,” she said, “but what customers are saying is that they can’t get a single view of ‘customer’. So we start with that, no matter the size of the organization, this is something that is evident across-the-board.”
It’s likely that such a discussion will ultimately involve a discussion of Metadata, or in business speak, Business Definitions and Business Rules, but that conversation needs to include people on both sides of the organizational fence – business people and technology people.
In the interview, Burbank discussed a use case that involved a business wanting to do specific Data Analytics on high net worth individuals. They had a huge amount of data collected through their various legacy data systems and new Big Data platform. Once they started to perform analytics on all of it they discovered the quality of their data was suspect: there were 16 versions of a client with a similar name, one was bankrupt, two had the exact same name, addresses were missing, etc.. “They had to get the little data right before they could get the Big Data right,” she commented.
Metadata and Communication are Key Components
“Business people have gotten Metadata for a while,” said Burbank. “They understand things like lineage and business glossaries, they may not understand the details of the technology stack, but they get the point.” But, they may not understand that making one change in their system could ultimately affect sales, marketing, and R&D systems. Explaining such issues to them is not difficult; it just takes clear communication lines that start with bringing together all stakeholders.
Metadata Management, with a solid business focus and implemented through constructive communication between business and IT stakeholders, can provide a solid foundation for any other Data Strategy projects moving forward. “When I talk with IT people I can use techy speak. When I talk with business people I speak in terms they understand,” she said. “But teaching others within and organization to do that is not always so simple. There are silos and territorial issues to deal with.”
Bringing both sides to a shared table, where the business challenge is openly discussed within business terms, while also making sure the technological implications are understood, can clear up many of those barriers. Data people understand that data can transform the business; business people understand the business processes, business needs, and want the data they use to be higher quality. Burbank observed, “I have spent a lot of my time helping determine what roles make sense. Business and IT people need to work together. Where does the overlap occur? It’s different for everyone, and in many cases a discussion of Metadata, no matter what terminology is used, can get that discussion started.”
Self-Service BI is driving a lot of this. The business is looking to take control of more of the reporting and even data integration to support it. If there is no Metadata, it’s difficult to understand the source systems. Metadata is the “glue” that helps bind business needs and technical implementations together.
Don’t Fall for the Hype
This discussion started by listing a few technologies such as Big Data, Internet of Things, and Machine Learning that are currently on the “all cool” list. But are they really? Does an organization need those technologies to gain greater market prominence? The answer may be “yes,” it could even be “certainly,” and “without a doubt.” Such technologies are changing the way business is being done globally, but what can they do for a given organization?
“There’s so much out there,” said Burbank. “There’s a lot of buzz, and there is a lot of need, and companies are looking for guidance, but before we move forward with a discussion of those different tools, we need to look at the business benefits. You don’t replace a functional Data Warehouse with Hadoop; they have different use cases all together, but it’s all very confusing to many people.”
They may need a new modeling tool, or a new MDM tool, or to discuss a Hadoop or Spark integration, or implementing Machine Learning algorithms for a new Predictive Analytics engine, or any number of other possibilities. Before such a discussion should ever happen, they need to focus on aligning their Data Strategy within their overall Business Strategy. Without proper preparation, all the newest technologies on the market will end up just costing more time and money. “People get caught up in the hype,” Burbank said. “But what is actually realistic? What are companies actually doing with this stuff? Is it worth it?” Such questions can be put on the table and, for many organizations, could be a central piece of the discussion, but not until the foundational elements of their data are in place – that takes planning and communication. “We start with the business drivers and go from there,” said Burbank. “You need to understand the business and speak their language, and then the rest can begin to fall into place.”