Poor data is a ubiquitous problem that plagues industries alike – from industrial manufacturers to property developers and construction companies.

Gartner surveys have revealed 59% of organizations do not measure data quality, while MIT Sloan estimates bad data costs companies anywhere from 15% to 25% of revenue. In construction, access to structured data is a sticky wicket – with data quality, fast-moving projects (where data inputs change quickly) with disconnected data sources, and data security among many issues that challenge the industry.

But organizations don't fail because they lack tools, or from a dearth of data; in fact, Gartner predicted as early as 2022 that 65% of B2B sales organizations will transition from intuition-based to data-driven decisions by 2026. It's not the tools; it's because they normalize low-quality inputs.

Data Is Only as Good as Its Input Quality (i.e., at the Source)

Data quality and quality control issues are everywhere.

Construction

The construction industry is no stranger to communication breakdowns and data management problems.

In the case of London's Crossrail project, poor data acquisition (i.e., poor coordination between different contractors and teams, resulting in inconsistent project data) led to costly delays (costing 24 billion pounds to complete, with the original estimate at 14 billion pounds). The Big Dig megaproject in Boston was also rife with problems that could have been prevented with better project coordination (and data management) – e.g., cost overruns, delays, leaks, design flaws, complaints for poor execution, etc.

Business

As Tom Krantz and Alexander Jonker for IBM discuss, the rise of AI further complicates data quality, introducing scalability concerns that need to be addressed, and of course humans need to be empowered, not undermined. Without proper governance, real-time monitoring, or data validation or remediation, AI hallucinations surge and further reputational hazards become inevitable.

Even in digital marketing, using customer data in an attempt to add personalization at scale can wreak havoc when segmentation and targeting are too unsophisticated. For example, a single transactional signal can mistakenly trigger intent: ordering that one business-adjacent product finds a misplaced contact flatfooted, in the middle of a long-term B2B nurture stream. In this case, the system didn't "fail" in the traditional sense; the inputs were too finite.

Industrial Manufacturing

In high-temperature manufacturing environments – such as cement, metal processing, glass production, or thermal treatment – many quality and safety decisions are still made using indirect signals, with operators inferring conditions from downstream defects, periodic sampling, or sensor proxies that don't fully represent what's happening in real time.

Possible Solutions

Construction

In construction, researchers have found data silos prevalent, with the lack of data standardization, lack of detailed categorization, and limited detailed information among the issues that have led to jobsite accidents; however, they have found that linking data systems and formalizing expertise and knowledge in safety reporting to be key anchor points to improving outcomes.

Furthermore, researchers have sought to understand how large data sets can be best used to help power pattern recognition, machine learning, visualization, and other data-driven executable insights for better outcomes.

To resolve many of the data quality issues that plague the construction industry, companies have also found good success with lean construction methodologies, wherein data quality can be improved by treating it as a production control strategy rather than an administrative exercise. Tools like pull planning and the Last Planner System depend on reliable construction data inputs — e.g., accurate lookahead schedules, current BIM models, verified material lead times, and up-to-date submittal logs. When version control is inconsistent or field reporting tools are not integrated with master schedules, teams make commitments based on incomplete information, and unintentionally introduce variability that undermines Percent Plan Complete (PPC) and disrupts flow.

Construction leaders can reduce this hidden waste by establishing clear data ownership across trades, enforcing standardized naming and coding conventions across platforms, and aligning digital updates with coordination cycles. In lean systems, information functions like inventory: when it is inaccurate, delayed, or fragmented, it creates rework and instability just as surely as excess materials on a jobsite. Furthermore, when companies move beyond the isolated project mindset toward a more industrialized process and philosophy, they can become full-scale, data-driven enterprises better positioned to scale.

Business

In the case of businesses looking to better nurture their customer audiences, segmenting your large email list can reduce potential liabilities; deeper categorization (e.g., industry + job title), along with intent signal thresholds can help ensure your messaging reaches the right audience at the right time.

Industrial Manufacturing

In the case of a challenging environment like a cement manufacturing plant, process control has been widely studied. Benefits have included implementing process improvements, like the Lean Six Sigma Method. Preventive maintenance also showed fruitful in decreasing the probability of failures and the number of consequent defects. What's more, manufacturers are increasingly focused on improving the fidelity of data captured at the point where material transformation actually occurs – rather than relying solely on assumptions or delayed feedback.