Enterprises are spending billions on artificial intelligence, cloud modernization, and analytics initiatives, yet many are building on foundations made of sand. In his latest book, The Cost of “Good Enough” Data: Why Modern Architectures Fail at Scale, veteran database strategist and technology author Craig S. Mullins delivers a candid wake-up call to enterprise leadership: the critical bottleneck in modern tech transformation is not the algorithms, but a chronic accumulation of data debt.
For decades, IT organizations treated imperfect data as an acceptable compromise. Broken pipelines, orphaned records, duplicate entries, and undocumented transformation logic were tolerated in the name of delivery speed. That calculation, Mullins argues, has shifted fundamentally.
“Artificial intelligence has changed the economics of data quality,” says Mullins. “When bad data feeds dashboards, the damage may be limited. When bad data feeds AI systems that make recommendations, automate processes, or influence decisions, the consequences can be amplified across the entire organization.”
His core thesis cuts through current industry hype: enterprises do not have an AI problem; they have a data problem that AI is actively exposing.
The Cost of “Good Enough” Data breaks down why massive capital deployments into data lakes, lakehouses, and distributed cloud platforms frequently fail to deliver expected business value. Instead of resolving foundational gaps, these complex hybrid environments often amplify architectural fragmentation, creating governance blind spots and compounding operational risk.
The book outlines several crucial disciplines required to navigate this landscape:
- The actual compounding cost of technical debt and degraded data quality across enterprise workflows.
- How distributed pipelines and architectural sprawl dilute organizational control.
- The direct link between verified data provenance and defensible, trustworthy AI systems.
- The strategic relevance of core systems of record, metadata, and active data lineage.
- The evolving operational scope for database administrators, data architects, and governance leads.
Rather than proposing another layer of platform software, the work focuses on structural, architectural, and organizational rigor. It serves as a practical blueprint for CIOs, CDOs, enterprise architects, and technical leaders who need to turn corporate data estates into dependable, long-term assets.
About the Author:
Craig Mullins is the President and Principal Consultant of Mullins Consulting Inc. An in-demand analyst, speaker, author, and practitioner, he brings over three decades of proven real-world experience across all facets of database systems development.
His background spans database administration, systems analysis and design, data modelling, performance management, and corporate technical education. Recognized globally for his deep expertise in enterprise information strategy and mainframe data infrastructure, Mullins continues to advise global organizations on data governance and architecture.
The Cost of “Good Enough” Data: Why Modern Architectures Fail at Scale is available now via Amazon books here: https://www.amazon.com/dp/B0H636956Y
Additional details, review copies, and speaking inquiries can be directed to Mullins Consulting, Inc. via craig@craigsmullins.com or mullinsconsulting.com.
