Quail Group: businesses fail with AI not because of weak models, but due to data confusion

Most companies struggle with AI not because of weak models or budget shortages. According to Quail Group's authors, the problem runs deeper: businesses don't understand which data truly matters, and so with the help of AI they simply scale their own confusion faster.

It's not about models

The market still has plenty of expectations that new investments will automatically bring more intelligence and efficiency. But the picture turns out differently. In an article for TNW, the authors reference State of Enterprise AI 2026: global AI spending could reach $2.52 trillion, yet only 14% of CFOs see measurable returns. Another alarming signal—42% of companies in 2025 abandoned most of their AI pilots. This looks not like isolated implementation errors, but like a systemic gap between ambitions and actual execution.

The authors argue against the popular explanation that everything boils down to "dirty" data. Cleanliness matters, but by itself it achieves little if the data isn't tied to specific decisions, isn't unified across systems, and isn't suitable for daily work. Many companies have accumulated dashboards, reports, and tracking systems that create an illusion of transparency. Meanwhile, teams often can't explain why a metric changed, how it affects results, or what action should follow at all.

How chaos grows

The problem is compounded by scale. Data volume grows faster than companies' ability to interpret it. Teams measure everything they can measure, but don't always understand why. As a result, dozens of metrics compete for attention, definitions diverge between departments, events are recorded differently, and reporting often depends on manual fixes. In such an environment, it's hard to assemble a unified business picture: everyone works with fragments, and fragments rarely align.

When AI is layered on top of such a foundation, confusion doesn't disappear—it spreads faster. Systems trained on contradictory input data don't eliminate ambiguity; they amplify it. According to data cited in the article, 61% of data leaders say that improving data quality helps move AI initiatives into production, but 50% still consider data quality and access serious barriers. Particularly concerning is the gap between confidence and understanding: 65% of leaders believe employees trust data for AI, while 75% simultaneously acknowledge gaps in data literacy skills.

"AI and automation amplify the condition of the data they rely on."

Where to start

The authors don't believe the problem will be solved simply with more convenient tools. If processes inside a company are initially unclear, metric owners aren't defined, and signals themselves are poorly documented, then any new AI system will operate over the same organizational fog. So they propose starting not with new models and not with another dashboard, but with rebuilding decision-making logic.

The practical start looks like this:

• Find the questions your business struggles most to answer today
• Assign owners for key data and metrics
• Standardize processes so events are recorded consistently
• Remove unnecessary indicators and keep signals tied to actions
• Build a cohesive data layer that's convenient for daily work

Special emphasis is placed on the human side. Even well-structured data won't help if the team doesn't understand how to apply it in daily decisions. So change management here isn't an optional add-on, but part of the AI strategy itself. Companies need to teach people to distinguish meaningful signals from background noise and act on them confidently, not simply consume more and more reporting.

What it means

The article's main point is simple: AI doesn't automatically cure organizational chaos. If a company lacks clarity in processes, accountability, and data, new models will only accelerate the output of questionable conclusions. Winners will be those who first bring order to their signals, and only then scale automation.