Data Management develops in stages across the enterprise, according to the DATAVERSITY® Trends in Data Management Report. One or a few teams take the lead doing Data Management, ensuring successful data knowledge, protection, access, and value for projects or products. Then the good practices stay with that team or project, while others miss out on this knowledge. This problem seems to be intensifying.
According to a Harvard Business Review article, the percentage of firms identifying themselves as being data-driven, using data as critical evidence to help inform and influence strategy, has declined in each of the past three years since 2017, suggesting poor Data Management. Firms need to have good Data Management to be data-driven throughout, and to take advantage of new technologies such as AI. The quicker good Data Management spreads, the faster a company becomes data driven.
Plenty of articles have general tips and tricks about doing Data Management in a company. However, coordinating other teams with past or current successful Data Management efforts can be very challenging. First, executives are not always aware of where successful projects exist in their organization and not all of their teams want to know. Take the story of Kodak. Sasson and Robert Hills successfully teamed up to invent the digital camera. Kodak’s marketing department rejected this project and its data, contributing to Kodak’s bankruptcy. Second, regulations like the GDPR make team members cautious about sharing any project-specific data with other teams or projects. Finally, data silos make it difficult to take successful Data Management practices from one team and integrate them into others. Executives facing a combination of all of these challenges can find it hard to distribute good Data Management used in one project to others.
Despite these difficulties, companies can leverage their Data Management successes. Enterprises can do a combination of four activities to spread good Data Management more quickly:
- Consider putting a Chief Data Officer (CDO) in charge of Enterprise Data Management
- Identify teams that successfully do Data Management
- Support Data Governance with authority
- Foster Data Management mentorship
Putting a Chief Data Officer in Charge of Data Management
First, consider putting a CDO in charge of Data Management across the entire organization, especially in a business. A CDO has a large picture of all the data in a company, including which inputs and outputs flow among different corporate teams and departments. This makes an enterprise-wide focus necessary. Without an executive in this role, Data Management becomes confusing. For example. IT and business do not know who needs to clean up bad data, because their Data Management activities overlap. Someone “needs to be a business strategist, advisor, Data Quality steward, and Data Management ambassador.” That person needs to be an executive.
In the past, the Chief Information Officer (CIO) has overseen Data Management, as some companies see it mainly as a technical solution. However, good Data Management spans more than just technical solutions to people, activities, and all kinds of processes. Good Data Management requires good business teams too; a CEMEX case study demonstrates this.
Furthermore, Mark Samuels at ZDNet explains, the “CIO remains the go-to executives for enterprise IT concerns.” In contrast, a CDO has strong communication and Data Strategy creation skills to get teams and all sorts of people to adopt digital technologies and work together sharing knowledge. CIOs and CDOs may work together, but they do not do the same role.
Likewise, expanding the role of a Chief Analytics Operator (CAO) to oversee Data Management does not cover everything. A CAO translates data-driven insights into data-driven actions, dealing mainly with how data is used. In addition to considering how data is used, a CDO also looks at how data is managed and governed. Teams that have successful Data Management consider all three of these points and benefit from a CDO. A CDO describes a resource dedicated to good Data Management across the organization, including ensuring that teams can learn good practices from others.
Identifying Teams Who Successfully Managed Data
Skilled CDOs know how to create a meaningful Data Strategy, aligned with the business strategy. These executives use a Data Strategy to make a company competitive and unified under a coherent Data Management vision. Data Strategy looks at the company broadly, casting a wide net over all sorts of Data Management team successes. For example, CEMEX case study mentioned above shows how “an end-to-end customer Data Strategy” helps to see the “whole business picture — from bringing in new customers through social media, digital campaigns, and other opportunities to retaining customers by enhancing the customer experience.” From this kind of viewpoint, CDOs can create templates of what good Data Management practices look like.
After identifying Data Management successes, CDOs need to look for good examples throughout the organization. Fostering transparency through project retrospectives may unearth Data Management achievements for a few data points. A more robust assessment, an enterprise-wide capability maturity model will not only provide organizational Data Management capabilities but, if customized to do so, will turn up Data Management successes.
A State of Arizona case study shows how a capability maturity model, from the CMMI Institute, assessed a decentralized departmental network in Data Management. From this assessment came a Data Steward training course for those at state agencies, on learning how to handle their data successfully. Arizona also implemented a state-wide Data Governance, from the CMMI assessment, enforcing what they had learned.
Supporting Data Governance with Authority
Spreading and supporting successful Data Management across an organization must include company-wide Data Governance, “a collection of practices and processes helping to ensure the formal management of data assets within an organization.” Data Governance “coordinates activities, not only to be legally compliant but also to pass along best practices from one team to another.” Here a CDO can come in and ensure a working Data Governance, and provide the authority needed to validate Data Governance policies. In the meantime, different teams find a safe space to communicate about their data needs.
Data Governance, supported by an authority, has benefited other companies. For example, Freddie Mac got stuck in bad Data Governance cycles. So even if a team had success with Data Management, others would not have to seriously consider adopting the practices. Freddie Mac changed Data Governance challenges around by enlisting “top-down support,” to get people on board with the good Data Management practices. With leadership pressure, departments became more enthusiastic to take on new Data Management practices. Having that top-down support, as from a CDO, makes a difference in spreading Data Management successes.
Data Management Mentorship with Successful Teams
A CDO, successful Data Management, and Data Governance with authority may not be enough or practical. New Data Management processes and procedures can be frustrating, and workers feel adrift. Setting up a Data Management mentorship, in a way that gels with the corporate culture, helps transfer good Data Management practices. If mentors have the time, they foster a positive mindset by listening to Data Management frustrations, validating them, and reframing them more helpfully toward being data-driven.
Mentors also would support the technical side of a mentee, making data systems and solutions less intimidating. Third, good mentors would identify Data Management successes and bring them to the forefront. This type of support would help foster a data-driven culture more quickly.
Should a person not have the time to be a hands-on mentor, they can co-sign. Tamay Shannon describes cosigning as amplifying someone’s voice. She says that a “person supports the other person publicly, providing a space where they feel like they can belong, and get their ideas out there.” Providing Data Management mentors, so that other data workers feel a strong sense of belonging, results in a 56 percent increase in job performance for that employee.
Most importantly, successful mentorships give a sense of belonging. Although good mentorship examples focus on data scientists, the practices could be extended to mentoring Data Managers. Jen Underwood, senior director of DataRobot, talks about companies starting to have “senior-level data scientists begin to mentor other team members and establish some of the foundations and best practices,” for AI projects in particular. Reshama Shaikh speaks about how she mentors others and recruits her mentees for help in meetup groups, engaging in mutually beneficial relationships. Her suggestions would also benefit those mentoring data managers, through shared mutual participation and belonging.
Good Data Management does not have to be a one-team or -project accomplishment. Excellent Data Management practices can spread throughout the organization. Companies who put a CDO in charge of enterprise data, identify teams successfully managing data, support Data Governance with authority, and include Data Management mentors will allow the best Data Management practices to be shared more efficiently, making the company data-driven that much faster.
Image used under license from Shutterstock.com