The gap no one can close: Why data still doesn’t lead to action.

While I was spending a (long) week at Adobe Summit and Google Next, I was having a lot of great conversations with digital leaders on the floor.

The energy was high. There’s a lot of curiosity. And there’s plenty about what AI, analytics, and agentic systems could unlock. But if there’s one thing that came up over and over again, it’s this:

Understanding what’s happening isn’t really the problem anymore. Most teams can get there (eventually).

The harder part is what comes next.

How do you actually move from data to action without it taking hours, days, or pulling in half the organization?

What struck me wasn’t a lack of data. It was how much work it still takes to do anything with it. Teams can see what’s happening. But turning that into a true insight, and then a decision, and finally into action, still feels heavier than it should.

And that gap isn’t just frustrating…it’s expensive. Every delay between insight and action is a missed opportunity to improve experience, recover revenue, or prevent impact.

The industry is moving fast, but not always forward.

There’s a real sense of momentum right now. Everywhere you turn, there’s a new idea, a new capability, a new promise about how fast things are about to move.

AI is everywhere. Agentic is quickly becoming the language people use to describe what’s next. And there’s genuine excitement behind it.

But at the same time, there’s a noticeable gap between that excitement and how clearly people can articulate what any of it actually means in practice.

I heard a lot of frustration around the lack of practical application.

There’s no shortage of big ideas, big potential, and big promises…especially from vendors. But that momentum tends to fall apart when it comes to applying any of it to real workflows. Not necessarily because the technology isn’t there, but because the transformation required is much bigger than just plugging in new software.

For large organizations, this is moving faster than anything they’ve dealt with before. Change management was already hard. Now it feels like it needs to happen constantly, not in phases.

There’s also a lack of clarity around what success even looks like.

What are we actually trying to achieve? How do we measure it? How do we implement something like this in a way that drives real adoption across teams?

Meanwhile, the core questions haven’t changed. Teams are still trying to understand what changed, why it changed, and what to do next.

But the path to those answers has become more complex than anyone expected. More tools and more data haven’t simplified things. In some cases, they’ve made it harder to move once you actually find something.

The blockers aren’t what most people think.

A lot of the conversation right now focuses on AI itself. What models can do, what they can’t do, what’s coming next.

But that’s not where most teams are getting stuck.

If anything, the uncomfortable truth is that AI isn’t the problem here.

What I heard repeatedly was much more familiar. Data that doesn’t fully connect. Systems that have been layered on top of each other over time. Processes that require too many approvals. Teams that are each working toward the same goal, but not always from the same starting point.

None of that is new. But it becomes much more visible when you try to move faster.

Because when something changes in the business, it’s rarely one team that needs to respond. It’s marketing, product, analytics, and engineering all trying to understand what’s happening from their own perspective, often in parallel, sometimes in conflict.

And that’s where the time goes.

Not in pulling the data, but in aligning on what it means.

A lot of organizations are experimenting with AI on top of this. Pilots are happening. There’s real progress in pockets.

But scaling that across the business and making it something teams rely on every day, that’s where things slow down again.

Because adding AI to a fragmented system doesn’t remove the fragmentation. It compounds it. This is also why so many "agentic" solutions fall short in practice.

The trust gap is now the biggest barrier

What’s also changed, and maybe more than anything else, is how people are reacting to all of this.

There’s still excitement, but it’s paired with a level of skepticism that feels new.

A lot of conversations came back to the same moment: seeing something impressive in a demo, and then immediately asking, “But will this actually work for us?”

That gap between the story and the reality is starting to feel wider, not smaller.

And a big part of that is because people have been burned before.

They’ve invested in platforms that promised transformation and didn’t deliver. They’ve gone through long implementation cycles that never quite translated into real adoption. They’ve seen demos that looked incredible, but didn’t hold up once applied to their own environment.

So now, there’s a much higher bar. Especially when it comes to agentic AI.

It’s not enough to show what something can do. People want to understand how it works. What data it’s using. How it reached a conclusion. Why they should trust it.

That expectation for transparency came up again and again. Not as a nice-to-have, but as a requirement.

At the same time, there’s still a very real hesitation around handing over control entirely.

The idea of systems operating across teams, making decisions or triggering actions automatically, is powerful. But it also introduces a level of risk that most teams aren’t comfortable with yet.

There’s still a strong preference for human oversight. For being able to validate, question, and understand before acting.

And that tension between speed and trust is something every team is trying to navigate.

The shift: From insight to action.

All of this points to a shift that feels subtle on the surface, but is actually pretty significant.

For a long time, the focus has been on getting better visibility. More data. More access. More ways to explore what’s happening.

And that work has largely been successful.

But visibility on its own doesn’t solve the problem anymore.

What teams are really asking for now is a way to move faster once they understand something. A way to go from “we see it” to “we’re doing something about it” without friction slowing everything down in between.

That changes the role of analytics in a pretty fundamental way.

It’s less about helping people find answers, and more about making sure the right answers show up when they’re needed. It’s less reactive, more proactive. Less individual analysis, more shared understanding across teams.

And this is where the idea of agentic starts to make sense—when it’s grounded in how teams actually work.

Not as a fully autonomous layer that replaces people. But as something that can quietly do the work of connecting signals, identifying what matters, and surfacing it in a way that’s immediately useful.

The real value isn’t in answering the question you already know to ask. It’s in identifying the things you didn’t realize were worth asking in the first place, and connecting that insight directly to what needs to happen next.

Because at this point, the next phase of analytics isn’t about more insight.

It’s about finally getting to a place where teams don’t have to go looking for it.

Where the right answers surface when something changes, and teams can act without waiting. Where self-service means understanding, not just access. Where bottlenecks no longer exist.

That’s the shift: from reacting to proactively knowing, and from dashboards to decisions.