New report: 1,000 Decision-Makers on AI, Confidence, and Trust

New research from TheyDo reveals a growing decision-confidence gap: the widening distance between how quickly AI helps enterprises make decisions and how confidently leaders feel about acting on its recommendations.

AI has delivered on one of its biggest promises: speed.

Of enterprise decision-makers responding to our survey, 68% say their organizations are making decisions faster since deploying AI. And the use of AI itself is already remarkably widespread. 94% regularly use tools such as ChatGPT, Copilot, or Claude for personal use at work, while 69% say AI is now embedded across multiple teams in their organization.

But faster decisions are not necessarily better or more confident decisions.

New research commissioned by TheyDo and conducted by Sapio Research among 1,000 business decision-makers suggests enterprises are running into a new constraint. AI can analyze information and generate recommendations almost instantly, but the people responsible for acting on those recommendations still need to understand what sits behind them.

And that context is proving much harder to access.

77% of decision-makers say their dashboards, CRM data, and customer feedback tell different stories at least some of the time.

That fragmentation creates a problem AI alone cannot solve:

The enterprise decision-confidence gap

For years, organizations have invested in collecting more customer and operational data. AI has dramatically increased their ability to process it.

Yet the information needed to make a decision still lives across different systems, functions, and teams.

One dashboard shows declining conversion. Customer feedback points to confusion during onboarding. Service data reveals increasing contact volumes. Product sees a feature adoption problem. Operations sees a process that appears to be performing efficiently.

All of those signals may be correct.

The difficult part is understanding how they relate.

Our research found that 43% of decision-makers, or roughly 4 in 10, delay or avoid making a decision at least once a week because they don't feel they have enough context about the customer or business impact.

The pressure is moving in the opposite direction. 77% say they are expected to make decisions faster than they can fully assess the consequences.

So enterprises find themselves caught between two realities: AI is dramatically reducing the time it takes to get an answer, while fragmented organizational context still limits the ability to decide whether that answer is the right one.

That is the decision-confidence gap. And it becomes particularly important at the top.

What’s wrong with C-suite AI decision-making? 

The people making the biggest decisions have a context problem

Executives are expected to see across functions, weigh competing priorities, and understand the consequences of decisions that can affect millions of customers.

Yet they are often further removed from the underlying customer reality than the teams closest to it.

Our research found that 27% of C-level respondents identify lack of context as their leading barrier to effective decision-making. 4 in 10 C-level leaders delay decisions several times a week. Across the full sample, the picture is more evenly split: slow or complex processes lead at 22%, with lack of visibility and lack of context close behind at 21% each. But process alone does not explain the challenge.

The bigger issue is that information can exist without being usable as context.

An executive can have access to dozens of dashboards and still not know why something is happening. They can receive an AI-generated recommendation yet still be unable to see the evidence supporting it. They can optimize a KPI without understanding how that change will affect the rest of the customer experience.

More information does not necessarily solve that problem.

In fact, organizations report using only 49% of the data and insights they collect to inform decisions.

The competitive question is shifting from How much data do we have? to Can we connect what we know well enough to decide what to do next?

Can local AI optimizations hurt the wider customer experience?

Yes. A decision that looks successful for one team can still create costs somewhere else in the journey.

The consequences become particularly visible when AI is used to optimize individual parts of a business. Imagine an AI-supported initiative that reduces average handling time in a contact center. On the team's dashboard, the intervention looks successful.

But what if customers now have to contact the company twice? Or an automated onboarding change increases completion rates but creates confusion later in the journey? Or a product intervention improves adoption for one customer segment while increasing friction for another?

Customers do not experience an enterprise one-department, one-system, or one-KPI at a time. They experience the cumulative effect of decisions made across all of them. That makes connected context essential.

Of the decision-makers we surveyed, 38% say their organization has already experienced an unintended customer experience issue or negative impact resulting from AI-driven decision-making.

Among those experiencing problems:

• 51% reported an increase in complaints or support queries.

• 46% experienced confusing or inconsistent customer experiences.

• 43% saw increased delays or friction.

• 42% experienced a loss of customer trust.

• 39% saw customers drop off before completing what they intended.

  ---

The problem is compounded by how organizations discover these issues.

Across customer experience problems more broadly, only 42% say issues are most commonly detected proactively, before customers are affected. Another 25% first notice a problem when operational metrics shift, and 30% don't find out until customers complain or churn. 

By then, the customer has already experienced the consequence.

For enterprise leaders, this creates a fundamental challenge. A decision can appear successful from the perspective of one function while creating costs, friction, or lost revenue somewhere else.

Understanding the impact requires seeing beyond an isolated process or transaction to the journeys it connects to.

Does trustworthy AI need more than data?

Yes, but volume isn't the same as context. 

More data alone doesn't automatically give leaders what they need to trust an AI recommendation. Customer feedback, operational metrics, behavioral data, research, and business KPIs each tell part of the story. What matters is understanding how those signals relate to one another, where they sit in the customer journey, and what else could be affected by acting on them.

The research shows just how important that visibility is to trust.

86% of decision-makers say they would trust AI-generated recommendations more if they could see and understand the underlying data.

And another 86% say they would trust AI recommendations more if they could see the customer journey behind them.

This is an important distinction.

Enterprise leaders do not simply want AI to give them an answer. They want the ability to understand why that answer makes sense.

In fact, 54% prefer to interrogate data themselves rather than rely solely on AI-generated summaries.

And when asked where they turn first when making important decisions, 38% choose their own experience, compared with 19% who turn first to dashboards and 17% to generative AI.

Human judgment is not disappearing from enterprise decision-making.

It is becoming more important because AI can produce recommendations faster than ever.

The role of technology should therefore be to strengthen that judgment by making the relevant evidence and context easier to understand.

The next stage of enterprise AI is shared context

Among organizations where AI is widely embedded across teams, 43% report fully integrated visibility of the customer journey. Where AI remains confined to isolated pockets, that falls to just 12%.

That means organizations with widely embedded AI are more than 3.5 times as likely to report full journey visibility.

The same pattern appears in decision confidence. 73% of respondents at organizations where AI is widely embedded report high confidence in decision-making, compared with 43% where AI remains isolated.

These findings do not prove that deploying AI more broadly leads to greater visibility or better decisions.

But they point toward something important: AI appears to become more valuable when it operates as part of a connected way of working.

That means the next stage of enterprise AI is not simply about deploying more sophisticated models or connecting those models to more databases.

It is about building an environment in which humans and AI can work from shared context.

How do you build the shared context that changes what AI can do?

Turn fragmented data into a connective structure AI can reason over through journeys.

A journey provides an organizing principle for fragmented enterprise data. Customer feedback can be connected to the moment it describes. Operational metrics can be understood in relation to the experience they affect. Research, behavior, business performance, and initiatives can be viewed together rather than interpreted independently. And critically, journeys can connect to other journeys.

That matters in a complex enterprise because decisions rarely affect a single experience in isolation. A change to onboarding may affect support. A service decision may affect retention. An intervention designed for one segment may create consequences for another. One journey can sit inside or intersect with several others.

Without those relationships, both humans and AI are reasoning from fragments. With them, it becomes possible to ask a much more useful set of questions:

What is happening, why is it happening, what else could be affected, and what should we do next?

This is the role of experience context. It gives people a shared view they can validate and gives AI a coherent structure to reason over.

Is having AI still a competitive advantage?

No. Decision quality will matter more.

As AI capabilities become increasingly accessible, simply having AI will become less of a competitive advantage.

At the same time, 52% say uncertainty prevents them from making necessary decisions at least sometimes. That suggests the opportunity for enterprises is not simply to eliminate uncertainty or automate more decisions. It is to create the infrastructure that allows people to understand uncertainty well enough to act.

That requires four things:

Connect customer context at the point of decision. Bring customer feedback, operational metrics, behavioral signals, and journey data into a living view of how customers, teams, and experiences relate to one another.

Make AI recommendations traceable. Decision-makers should be able to see the evidence behind a recommendation, question it, and validate it before acting.

Assess impact across the wider customer journey. A decision should not be evaluated only against the KPI or process it was designed to improve. Leaders need visibility into the consequences for other journeys, teams, segments and stages of the customer relationship.

Create shared ownership across functions. Customer experience is produced across the enterprise. The context used to make decisions about it also needs to travel across the enterprise.

Most large organizations already have much of the underlying information. What they lack is the connective layer.

The question is no longer whether your enterprise has enough data

It's whether your enterprise has enough context.

AI has made generating answers easier than ever. The harder problem is knowing whether an answer deserves to become a decision. That requires something no model can create from fragmented information alone: context.

For enterprise leaders, the next phase of AI adoption therefore demands a different question. Not simply, How do we use more AI? But: What does our AI understand about our customers, our business, and the consequences of the decisions we're asking it to help us make?

The organizations that can answer that question will be in a fundamentally stronger position. Because when AI and people share the same connected customer context, speed is no longer the objective on its own. It becomes possible to move quickly and understand why you are moving.

That is the foundation for better decisions. And as AI itself becomes ubiquitous, better decisions may prove to be the advantage that matters most.

About the research

TheyDo commissioned Sapio Research to survey 1,000 business decision-makers with direct influence over budget allocation at organizations with more than 1,000 employees. Respondents represented financial services, utilities, retail, health care, manufacturing, and telecommunications across Belgium, Germany, the Netherlands, the United Kingdom, and the United States. The research was conducted online in June 2026.

TheyDo is the experience context platform for enterprise teams. It connects customer feedback, operational data, and journey insights into a shared context, giving people and AI the evidence they need to understand what is happening, why it is happening, and what to do next.