The key to good data management? Start with good data

Editor’s note: Procurement Pulse is a monthly look at a topic or trend that is impacting the procurement function at organizations of all sizes. If you are interested in future topics, you can see a full list of upcoming features on our Editorial Calendar.

The increasing use of artificial intelligence and Generative AI in supply chain will not fix the primary problem that has existed since the dawn of time: data quality. The old saying, garbage in, garbage out, still applies to AI. But, paraphrasing the words of Yossi Sheffi, director of MIT’s Master of Supply Chain Management degree and a leading voice in supply chain: All AI is doing is allowing us to fail faster.

“In today’s competitive business environment, effective data management is essential for procurement success,” explains Niraj Chatwal, data strategy and AI leader for MorganFranklin Consulting. “By ensuring real-time data access, maintaining accuracy, and optimizing supplier relationships, organizations can enhance efficiency and make informed decisions.”

Good data management, though, starts with good data.

“Data management has bedeviled large companies for decades. Almost all firms spend a lot on it but find the results unsatisfactory. While the issue does not appear to be growing worse, resolving it is increasingly urgent as managers and companies strive to become more data driven, leverage advanced analytics and artificial intelligence, and compete with data,” wrote Tom Davenport, Theodoros Evgeniou and Thomas C. Redman in a June 2021 Harvard Business Review article.

They could have written the same thing in 2024 as not much has changed. We simply have more data to manage.

The challenges to good data management

The challenges to collecting good data are not a secret: rapidly increasing data volumes, data accuracy and reliability, poor data governance, legacy systems and processes, and IT talent attrition are among the greatest challenges.

Let’s start with the last one: IT talent attrition. In a 2022 report, data company Hakkoda reported that only 3% of companies across all industries had no problem finding talent. Companies are particularly struggling to find machine learning-experienced data scientists (48%) and data architects (34%).

“It’s interesting to note that when asked which roles provided the most business value, data analysts (41%) and data architects (35%) topped the list. Data scientists were a close third,” the company reports.

Finding talent is likely to remain a challenge for the foreseeable future. The other challenges might actually be easier to overcome.

“Robust analytics capabilities help assess compliance rates and identify cost-saving opportunities by quickly identifying unnecessary [procurement] payments,” says Chatwal. “Additionally, addressing new and emerging ESG requirements and integrating AI-driven audit solutions ensure regulatory adherence. Prioritizing data management empowers procurement leaders to navigate challenges and elevate their impact within the supply chain.”

If only it was that simple.

Data volumes are going to continue to increase—estimates are that there are 402.74 million terabytes of new data being generated each day. That won’t change, so companies simply need to have the systems in place to accommodate and analyze all the data they will generate.

Data accuracy and reliability

Putting in place good data governance is a key to good data management.

“Traditional data governance tools built for data warehousing prove more of an obstacle than an assist in today’s environment,” Hakkoda wrote in a blog posting on the topic. “Data can be sourced from anywhere—and as such, siloing this information creates more problems than it solves. Companies need data governance tools that work with these new frameworks. To combat these problems, organizations must establish a coherent, limber data governance framework that centers on ever-updating compliance standards and guidelines.”

Notably, Hakkoda said that a good data governance framework is about the people, not the processes.

“All the technical wizardry in the world can’t implement a data governance program if the political will isn’t there for the organization to put in place a framework and process for governance,” it said.

Assuming companies can get the data governance part down, are able to work around an IT talent shortage, and are able to upgrade legacy systems, the real work comes in how to address these challenges.

Chatwal offers three suggestions to procurement managers to manage data. It includes ensuring data accuracy and reliability, focus on business functions and processes, and enable better analytics with clean data.

“Consistently validate and refine your vendor master data,” Chatwal says. “Accurate data minimizes the risk of errors, such as duplicate payments. Clearly defined data elements empower your team, delineate roles, and strengthen overall accuracy.”

Chatwal goes on to say that by identifying key business functions and processes, businesses can “gain insights into departmental needs, existing challenges, and pain points [and] understand the technology landscape—what tools are in place to manage data—and align efforts with overall business objectives.”

Finally, as mentioned in the opening, good data is the key. A quality data management program is not possible without clean data.

“High-quality vendor data serves as the foundation for robust analytics,” Chatwal says. “Spend analytics, for instance, relies on accurate and comprehensive data to uncover patterns, identify cost-saving opportunities, and optimize procurement strategies. Clean data ensures your spend analysis is both reliable and actionable.”

SC
MR