Analyst Perspectives

If a single phrase could sum up the big data craze of a dozen or so years ago, it would be “more data beats better algorithms.” Attributed to Google research director Peter Norvig, the quote effectively summarized a research paper Norvig jointly authored called The Unreasonable Effectiveness of Data and was embraced by big data enthusiasts as articulating the prevalent thinking that enterprises with the largest volumes of data have an advantage over rivals. The phrase was, of course, an oversimplification, and enterprises investing in big data projects quickly found that quantity was not the only characteristic of data that mattered.

An adapted version of the quote evolved: “More data beats clever algorithms, but better data beats more data.” Today, this is also often attributed to Norvig, although if my memory serves me correctly, the qualifying caveat should actually be ascribed to former LinkedIn data scientist Monica Rogati. Either way, the importance of better data is once again highly relevant as enterprises seeking to accelerate large-scale strategic adoption of artificial intelligence have recognized the critical importance of data management.

As I previously explained, while data is integral to AI, poor data management can be an impediment. Large volumes of data are required to train models, while data freshness is important to inferencing in interactive applications, and data quality is fundamental to ensuring that the output of agentic and generative AI initiatives can be relied upon. While AI-ready data is clean, well-organized and compliant with regulatory standards, too many enterprises struggle with data that is fragmented, inconsistent and not easily accessible. More than one-half (54%) of participants in ISG’s 2025 Market Lens Data and AI Program Study cited the usability of data for AI applications as a significant data challenge. As such, even if enterprises have proven the value of AI with small-scale initiatives, many have identified the need to take one step back by pausing to improve data management with a view to subsequently taking two steps forward with accelerated strategic AI initiatives.

ISG defines data management as the administration of data throughout its lifecycle, from generation to consumption. Data management has been a critical component of enterprise IT strategies for decades, enabling users to ensure that data is valid and consistent and can be trusted for operational use cases and analytic decision-making. Data management combines functionality addressing data governance, data quality, master data management, data integration and data intelligence to ensure that the enterprise is collecting, storing and processing data in accordance with strategic goals and regulatory requirements.

Data integration is a set of processes and technologies that enable enterprises to extract, combine, transform and process data from multiple internal and external data platforms and applications to maximize the value of analytic and operational use. Without data integration, business data would be trapped in the applications and systems where it was generated. Analysis of individual data sources—customer or product data, for example—can provide insights to improve operational efficiency. However, the combination of data from multiple sources enables enterprises to innovate, improving customer experience and revenue generation, for example, by targeting the most lucrative customers with offers to adopt the latest product.

Data governance enables organizations to ensure data is cataloged, trusted and protected, improving business processes to accelerate analytics initiatives while supporting compliance with data privacy and security policies as well as regulatory requirements. While not all data governance initiatives are driven by regulatory compliance, the risk of falling afoul of privacy (and human rights) laws ensures that regulatory compliance influences data-processing requirements and all data governance projects. Good data governance provides guardrails that enable enterprises to act quickly while protecting the business from risks related to regulatory requirements, data-quality issues and data-reliability concerns.

Maintaining data quality and trust is a perennial data-management challenge, often preventing enterprises from operating at the speed of business. As enterprises aspire to be more data-driven, trust in the data used to make decisions becomes more critical. Without data quality processes and tools, enterprises may make decisions based on old, incomplete, incorrect or poorly organized data. The precise measure of quality will depend on the individual use case, but important characteristics include accuracy, completeness, consistency, timeliness and validity.

Creating a “single version of the truth” that provides an agreed definition of customers, products, suppliers or workers is a perpetual challenge for many enterprises. MDM is the practice of establishing and protecting foundational reference data used by an enterprise to provide an agreed list of entities that can be shared throughout the organization. MDM encompasses data validation, matching and merging duplicate records and enriching data with related information, as well as data modeling, which documents the relationships between data elements.

Data intelligence provides a holistic view of data production and consumption, enabling data administrators to understand and manage the use of data in business intelligence and AI initiatives and accelerate strategic data-democratization initiatives to provide data analysts and business users with governed, self-service access to data across an enterprise. The term data intelligence has been used by multiple software providers across analytics and data for several years. Within the last 12 months, it has become a more clearly defined product category. As such, we will be taking a more focused view of data intelligence in our forthcoming ISG Buyers Guide, compared to the broader definition we used in the 2024 Data Intelligence Buyers Guide.

ISG defines data intelligence platforms as providing a combination of data inventory, data discovery and metadata management functionality, as well as data governance, data quality and data lineage. Data intelligence platforms ensure that business users and data analysts can find and access the data they need while providing analytics and data leaders with key metrics on data production and consumption, including the value generated by data projects. I assert that through 2027, enterprises will prioritize data intelligence software providers capable of providing a holistic view of data production and data consumption across the organization.

Along with overall data management, our forthcoming ISG Buyers Guides will address these functional areas with dedicated reports focused on data integration, data governance, data quality, MDM and data intelligence. Enterprise use of AI is increasingly a board-level concern, raising expectations in relation to efficiency, innovation and growth improvements. I recommend that all enterprises exploring the use of AI invest in refreshed approaches to data management to ensure that the data used for AI initiatives is fit for that purpose.

Regards,

Matt Aslett