In a recent panel hosted by The Doyle Group in their online community, the briDGe, industry experts delved into a range of topics that are shaping the future of data management. This panel was the second session in the Data Insights series. The session built on previously discussed topics, including how to integrate data strategies at a project’s inception, how to identify key qualities in top-tier data professionals, data governance best practices, and more. To access part one of DG’s Data Insights series, please click here.

Offering their perspectives in this conversation were two influential leaders in the industry:

• Cher Fox, President and Founder of Fox Consulting, International Speaker, DAMA Board Member
• Karen Pfeifer, Data Strategy and Data Product Executive Leader at Audacy

Doyle Group Founder Andrew Doyle served as the discussion moderator. Here are the topics discussed and expertise shared during this session:

What strategies are most effective in optimizing data storage and infrastructure to support scalability and performance?

In terms of data storage and infrastructure, data engineering teams can often focus on “low-hanging fruit” first. This includes cleaning up old “tech debt” from a company’s system. (Think legacy databases and their platforms, temp tables, and duplicates, for example.)

One helpful rule of thumb is the 70/30 Rule. Data engineers should spend about 70% of their time on new work, and 30% on cleaning up tech debt and other maintenance tasks. Over time, the tech debt will be “paid off.” As a result, teams can focus even more on growth-oriented tasks for the organization.

It’s also important to never discount the value of DBAs (Database Administrators). Developers tend to come at data storage systems from the perspective of “What can we make this do?” In contrast, DBAs often ask questions like “What do we need this system to do?” and “How can we effectively manage this dataset day in and day out?”

“We’re doing it on a continuous basis. It’s not just some quarterly review, it’s a constant review. I believe you really can’t stay ahead of it otherwise.”

-Karen, on how frequently companies should assess their data optimization strategies

What are some of the key strategies for ensuring data quality and consistency across teams?

There are several key elements that contribute to the consistent collection, use, and storage of high-quality data. For instance, a company needs to define its data quality standards and complete data profiling. Also, data governance can be a huge factor in delivering exceptional data quality across departments.

There may be some trepidation around limiting or restricting data collection and/or usage. However, judicious governance can help key stakeholders screen out the “noise” from large data sets and focus on the information that’s truly important for continued growth.

In addition, it’s vital to train the humans who input and work with data on a daily basis. Company leadership needs to make clear that data quality is a top priority. Collaborative processes should be put in place. This will make it easier for cross-functional teams to generate accurate data, store it effectively, and even interpret it in ways that are understandable across different departments.

“Data ownership isn’t just for data teams, it’s not for data leaders. Everyone in an organization is responsible for data: if they use it, if they consume it, if they create it, if they manipulate it. And the more people that are… participating in that party, hopefully the cleaner the data is and the more value it is generating for the organization.”

-Cher, on the importance of taking ownership of company data across departments

What are some data quality metrics and key performance indicators to benchmark and assess effectiveness?

This really depends on the nature and needs of the organization. If a company is “drinking from the fire hose” in terms of taking in large data sets, it will have to pick and choose its battles. It will need to decide which KPIs are most directly correlated with organizational growth.

As another example, a small startup business may seek accuracy, completeness, and validity for its data sets. However, it may not have an urgent need for accessibility right now. Of course, data error and correction rates are almost always key metrics, since they relate to the concept of data trust. 

“If people are hesitant to use the data, then somewhere in the back of their mind they don’t trust it, and it’s not providing value. And they find too many errors, and they spend too much of their workflow trying to clean that up, and trying to reconcile… And really beautiful analytics products will be abandoned if the users don’t trust it… That’s a huge waste of resources, money, and a variety of other things for organizations.”

-Cher, on how lack of data trust can negatively impact companies

What are the key factors that determine successful AI adoption and integration?

Most large companies are already using AI in one form or another. The newer, generative AI technologies on the market (natural language processors, advanced machine learning algorithms, etc.) may be useful for automating or enhancing certain tasks, but the education of leaders and team members needs to come first. Executives should ask: “Will AI fill a specific need in our business process, and if so, how?”

Once the decision is made to adopt AI, leaders should also be aware that many machine learning models have already been built and produced for public consumption. This means that management needs to decide whether to build an AI model from scratch, buy a model, or use an open-source platform.

“Integrating AI for decision-making starts with understanding what AI is, and then it starts with understanding the differences of applying tools that are already out there versus trying to build something new — and even then, [leveraging] models that are already out there… for purpose-built adoptions for your company.”

-Karen, on how organizations can effectively integrate AI into their decision-making processes

If you’d like to learn more about why data quality, consistency, and governance are so important, or how to successfully integrate AI into your business processes, access the full panel discussion here. Moreover, if you’d like help finding the right data consultants and professionals for your company’s needs, reach out to our team at The Doyle Group today.