Here’s a scenario that might feel painfully familiar.
Your marketing department captures customer leads, and passes them to the sales department. Marketing’s success is measured in part on the number and size of deals that result. But a squabble breaks out over how the sales department handles, nurtures, and attributes those conversions.
Result: Neither department really wants to share their data.
Logistics versus customer service, IT versus lines-of-business, one business unit versus another, and so on—conflicts over data ownership and data sharing can arise in any corner of the enterprise. In fact, control issues may be more the rule than the exception. “Data ownership is a struggle for most organizations,” said Rich Peters, founder of data consulting firm Tongere Partners.
Part of the problem stems from the proliferation of siloed data sources in many companies. The IDC State of the CDO research in December 2021 (commissioned by Snowflake elite partner Informatica) found that 79% of organizations are using more than 100 data sources, with 30% using more than 1,000 sources. To illustrate the effect, another survey conducted by Adobe found the top three obstacles to creating a single view of the customer were
- Integrating data sources
- Data silos
- Departmental silos
Couple that fragmentation with a human tendency toward protectiveness that can spill over into territorialism, and you’ve got problems. The potential value of data sharing across industries is well documented, but many companies are still stuck on the issue of internal sharing.
Here, experts share insights on overcoming control issues to foster data sharing within your company.
Ownership versus stewardship versus access
A valuable idea, which admittedly requires steady educational effort in most companies, is to separate ownership, stewardship, and access.
Every department can still “own” the data it captures or creates, noted Jennifer Belissent, Principal Data Strategist at Snowflake. As Belissent explained in What Makes a Great Analytics Roadmap, each group has the best understanding of the quality requirements for that data, and many ways that the data should or shouldn’t be used. This is often best handled by appointing a data steward to help set and ensure access and usage policies, while “coordinating that effort with broader enterprise governance policy you’ve set in place,” she said.
Peters agreed that data ownership and data access are two different things. “Someone in the organization is the natural owner for a piece of business or data,” he said. Visibility into how other groups want to use the data can help deter territorialism from taking root.
Consider data-sharing contracts
Homeowners know that if you hire someone to remodel a bathroom, but don’t get the agreement in writing, you’re potentially in for a world of hurt. A good contract provides all parties with clarity and assurance.
The same idea applies to data sharing. Belissent stressed the value of formal data-sharing agreements, even within a single organization. A data-sharing agreement documents the understanding and expectations described in the section above.
These internal agreements will also come in handy as your data partnerships expand to include more and more outside entities. The terms will change, and the legal department’s role in the process will surely expand, but the structure and internal understanding of how such agreements work, and the ways they don’t work, can make external agreements quicker and easier to execute.
Carrots are tasty (but don’t throw away your stick)
The idea of a “formal agreement” might meet with initial resistance in some quarters, so it’s vital to stress the benefits as much as the requirements.
This is true across many aspects of data sharing: “Positive incentives to participate work better than demands,” said Belissent. As an insights team grows, or as the enterprise data program moves along its roadmap, emphasize top enterprise priorities and constantly communicate about quick wins and business impact.
Transparent processes also help, so departments and business units can understand how their data is being used, where it is made available, and how it is contributing to increased business value. They will also appreciate proactive communication about new data catalogs or new reports becoming available for their own use, demonstrating reciprocity from other teams.
All of this work helps cement the idea that the insights program will help each participating group deliver on its own specific goals. Then, as governance requirements evolve and or new requirements land—and they inevitably will as the program matures—these groups will be less likely to fight against it.
Relentlessly build IT and business department engagement
From our first days on the elementary school playground, we learn that sharing is built on trust. That early lesson still applies to data sharing. Analytics programs benefit from constant, intentional relationship-building among all parties involved, experts noted.
Building a common language and understanding creates a foundation for collaboration. IT should pay particular attention to demonstrating—mostly by listening—a full grasp of business leaders’ priorities and challenges.
“Often you can take a whiteboard and hash out a lot of these definitional and ownership issues in an afternoon. Do you care about subsidiaries? How do you sell to them? What’s your funnel look like? How long is the sales cycle? What do you need to track?” said Peters.
“If you don’t understand what the organization does, how can you understand the data? Sometimes IT doesn’t have all the exact language, but when you ask about business processes, they can talk all day,” he said.
The quest for trust and a common language for IT and business is nothing new. Happily, that means that progress made in the name of analytics will certainly pay off in other areas of applying technology to business challenges. And in data sharing alone, the payoff can range from consistent reporting to more accurate models, deeper customer insights, and new market opportunities.