Data Summit Connect 2021 Second Day Keynote Marries Data Governance with Making Data Driven Decisions

The second day of Data Summit Connect 2021 opened up with presentations by Seth Earley, founder and CEO, Earley Information Science and author, “The AI-Powered Enterprise,” and Kevin Kline, head geek, SolarWinds, who discussed the importance of data and how to governance can help organizations drive decisions with data.

Early’s presentation, “Creating Data-Driven Decision Frameworks to Increase Speed and Quality of Decisions” focused on using data to drive decision making.

This means setting up scorecards and dashboards that will identify problems and guide changes—a true data driven decision framework.

This approach can be extended to many types of processes and systems once it is in place but getting stakeholders to trust the data can be a big stumbling block. Hard data can lead to better results—the ROI every leader is looking for.

“It’s difficult to drive decisions with data instead of an opinion,” Earley said. “We want to take that opinion out and drive decisions with data.”

When building a data decision framework, organizations need to measure, make changes and measure again, he noted.

Collect baseline master and interactive data that goes across multiple processes. This technique can be used for any process in the organization where you can collect a baseline.

Then organizations need to hypothesize the impact of design changes. It’s not just about the operational piece but it’s about the things that support that, he explained.

Organizations need to measure the impact of design chances with behavioral testing. Then course correct by identifying and implementing corrective and preventative measures.

“We need to understand where does the data come from and what it’s doing,” Earley said. “It comes from a fairly complex digital machine.”

Mapping those data sources, usages, ownership, quality, and more need to be articulated, he explained.

“If the data is important you need to spend the money on it,” Earley said.

Many times there is no correlation or connecting tools, mechanisms, or integrated decision making within an enterprise.

There are five data and content related domains, he added. This includes UX and usability, search, taxonomy, content and metadata, and product information. The simplest metrics are often the most important for meaningful benchmarking.

He offered several tips for product data metrics rollout:

  • Start simple and use what you know
  • Consider market references
  • Consider home-grown solutions and move to specialized software
  • Integrate with existing or emerging data governance programs
  • Use revenue in addition to item counts
  • Tie metrics to conversion rates
  • Analyze correlations between types of metrics (design, testing, web analytics)

“You can really apply this to anything,” Earley said. “Whatever your metric is you are also trying to understand the value of that data.”

According Earley, metrics programs should:

  • Use consistent definitions and explanation of impact
  • Communicate in plain language and summarized visuals
  • Separate business rules from those that are clearly required by specific systems
  • Involve data owners with vested interest in early wins

“You can really instrument any type of workflow in the organization,” Earley said. “You don’t want to overwhelm people. You have to separate the signal from the noise.”

Scaling integrated metrics programs involves regular collaboration across different data and content teams.

The Importance of Data Governance

Kline’s presentation, “Data Governance for DBAs” focused on the importance of high quality data. As organizations grow in complexity and sophistication, they begin to realize that their data is a true asset of the organization which, in and of itself, has enormous value that should be managed proactively.

Enter data governance, a set of concepts centered on ensuring that an organization has high-quality data throughout its life cycle.

A study by McKensey shows that poorly implemented data governance hurts data lakes, data integration projects, customized applications, and databases and apps with multiple SaaS integration points, Kline said.

Companies without a governance plan cannot find the right information, can’t combine or manipulate information, can’t access or consume information, can’t extract value or decision-making insight from information, and can’t protect data according to its true value.

“When you have data governance in place, organizations use the data to drive decision making,” Kline said.

Data governance architecture includes the measured business impact of data governance, information lifecycle management, master data management, data analytics, metadata management, data quality, and data privacy, security, and compliance.

Good data governance requires shared responsibility of tech and business, Kline said.

DBAs will be able to reuse data from existing resources, improve standards and use of data types. The architecture is more consistent and cohesive.

More information about Data Summit Connect 2021 is available here.
Replays of all Data Summit Connect 2021 sessions will be available to registered attendees for a limited time and many presenters are also making their slide decks available.