2023 State of Analytics Report with Jim Kobielus
James Kobielus, TDWI’s senior director of research for data management, explores the state of analytics in 2023, following the publication of his recent TDWI report.
- By Upside Staff
- November 16, 2023
In this recent “Speaking of Data” podcast, TDWI’s James Kobielus, explores the state of analytics in 2023 following the publication of his new TDWI report. Kobielus is senior research director for data management at TDWI. [Editor’s note: Speaker quotations have been edited for length and clarity.]
Kobielus began with a discussion of what is meant by “analytics maturity.”
“Essentially, maturity refers to how organizations institute practices, platforms, and skills to maximize the value of their data and analytics resources,” he said. To qualify as “mature,” such practices, platforms, and skills must also be consistent, stable, and adopted as standard practice throughout the organization.
When asked about how central analytics is to the modern business, Kobielus was keen to point out “running your business on data means running your business on analytics. That analytics may take the form of reports and dashboards, predictive models, AI -- what have you -- and making it must be available to everyone who needs it.”
Kobielus referred to the TDWI Analytics Maturity Model Assessment as the source for much of the current information about where organizations stand on their analytics maturity journeys.
“The assessment looks at several different dimensions,” he said. “First, organizational commitment -- are your leaders totally on board with analytics as a key business tool? Have they built a culture of data and analytics literacy around which they are building teams and embedding processes into the daily flow of work?” When discussing the challenges organizations are facing, Kobielus said this culture of data is often the biggest hurdle.
“Second is resource commitment. Are your leaders dedicating the human and financial resources necessary to build and manage the data, models, and applications required?
“Third is the data infrastructure. Has your organization adopted an elastic, unified cloud data platform to carry out much of its work? To what extent has analytics been embedded in your company’s operations?
“Finally, do you have a strong governance program in place to put controls on all your data and analytics?” Kobielus explained that say governance is especially critical to what he called “responsible data and analytics” -- the maintenance of data and analytics that are high quality, accurate, secure, and free of bias. Unfortunately, he noted, only about half of organizations have implemented model governance as a way to ensure their advanced analytics and AI models remain accurate and fit for purpose.
“So far,” he noted, “the assessment shows that organizations are scoring an average of 55 out of 100, or what we call the ‘Early Established’ phase of maturity.”
He also discussed several key modernization trends that have come up in TDWI research.
“One is the implementation of the cloud data lakehouse -- a combination of a data warehouse for structured data (for typical BI and analytics) and a data lake (for AI and machine learning),” he said. “Another is the development of low- or no-code tools for non-technical users to build their own analytics.”
Kobielus continued, “We’re also seeing a lot of interest in end-to-end life cycle automation to simplify the ingestion, cleansing, consolidation, and transformation of data to make it suitable for analytics.”
Another trend he mentioned was that of companies increasing use of data catalogs to help data scientists and business analysts alike search for and find what they need from the vast stores of data and analytics models that companies maintain today. “Providing employees with the tools to find and access data as easily as they might grab a box of paper clips out of the office supply closet is key to building the data-centered culture you’ll need to succeed.”
“Of course, there’s generative AI,” he said. “Organizations are still trying to grasp what tools are necessary for developers and data scientists to work properly with generative AI -- the kinds of modifications that will need to be made to the lakehouse, what changes will need to happen with model governance, and so on.”
Kobielus concluded the discussion by pointing out how few organizations are actually measuring the impact of their analytics programs.
“According to our research,” he said, “only about a quarter of organizations are measuring the quantitative bottom-line business impact of their analytics. This is obviously a problem because there’s no way to know whether you’re getting your money’s worth from your analytics efforts if you’re not measuring either top-line revenue generation or bottom-line cost savings. The real end state of analytics maturity is for it to be self-perpetuating and self-funding.”
Learn more in our TDWI State of reports.