If you’ve been trying to figure out why AI isn’t returning what you invested in it, the answer may be closer than you think. While some leaders may blame the tools, the more accurate root cause lies in two organizational conditions that either make or break AI value.
Those conditions are friction and enablement. They’re independent of which AI tools you’ve deployed, how much you’ve spent, or how enthusiastic your team is about using them. Together, they determine whether AI becomes a true organizational advantage or a wasted investment.
In this blog, we’ll break down exactly what each condition is and introduce the AI Value Quadrant, a framework from our latest research that maps where your organization sits across both.
The two variables AI value hinges on
The data across our report points to two organizational conditions that determine whether AI delivers returns or prevents them.
The first is friction. Friction is the overhead technical workers carry just to keep work moving across tools and systems. It’s the cognitive load that workers take on as they figure out which tool to use for which task. It’s the burden that comes from constantly what you’re working on each time a new tool is introduced. And it’s the effort that accumulates during coordination, before any actual building happens.
The impact of friction shows in the data. In high-friction environments, where tools are fragmented and work doesn’t flow cleanly across systems, 50% of technical workers say AI increases their workload. In low-friction environments, more than half say it lightens it. Same tools. Different environments. Opposite outcomes.
That flip is worth understanding. AI doesn’t operate in isolation. It inherits whatever environment it’s dropped into. When that environment requires constant context reconstruction and tool switching to get anything done, AI becomes one more variable to manage rather than something that manages things for you.
The second condition is enablement. Enablement is how well your organization has equipped workers to use the AI tools you’ve already deployed—not just by making them aware of what exists, but by giving them the training and capability to use those tools effectively in their day-to-day work. When awareness is low, training inconsistent, or capabilities too complex to build fluency around, adoption stays shallow, and value never fully materializes.
When friction or enablement is missing, AI value doesn’t reach its potential. When both are broken, value barely shows up.
The four states of AI value
Together, friction and enablement form the AI Value Quadrant, a framework from our latest research that maps where your organization sits across both conditions. The quadrant plots organizations along two axes: high or low friction and high or low enablement.
High friction is a liability. It means more overhead, more context loss, and more difficulty coordinating before any building happens.
Low friction means work flows cleanly across tools and
context
doesn’t need to be repeated.
High enablement is an asset; it means workers are equipped, fluent, and getting real value from the tools they have.
Low enablement means capability is sitting on the table, unused.
The combination of where your organization falls on each axis produces four distinct states, each with its own failure mode, signals, and path forward. Let’s dive in to each:
Stalled: High friction + low enablement
When friction is high and enablement is low, AI value stalls. This is where the gap between your investment and your return is the widest. It’s also the state where it’s easiest to misdiagnose what’s actually wrong.
In a stalled environment, your tool stack is creating friction that slows work down, and your teams haven’t been equipped with the guidance or training to push through it. Workers don’t know which tools to use or when, and even when they do, the tools don’t work together well enough to make it worth the effort.
New tools get rolled out without training, adoption plateaus early, and the workers who do get value are the ones who put in the extra effort themselves—reading the docs, watching the tutorials, figuring it out on their own.
For everyone else, when the official stack feels like extra work, they stop using the tools their company provides for them. In high-friction environments, 89% of technical workers report turning to AI tools their company didn’t approve. Ninety-six percent say they’d rather use a familiar tool even when they know a better one exists.
The result is AI value that stalls before it starts. Teams in this state are often the busiest, and they’re rarely the most efficient.
Constrained: High friction + high enablement
When friction is high but enablement is strong, AI value gets constrained. This state is one of the most frustrating places to be because your organization is putting in the work. You train your teams. You roll out new tools with adequate guidance. People know what’s available and how to use it. And yet the returns aren’t showing up at the scale you’d expect.
The problem is that the tools you invested in and trained your teams on are deployed in a high-friction environment. Context doesn’t carry between systems, so work has to be repeated at every handoff. Tools don’t talk to each other, so teams spend time bridging gaps that should close automatically. Decisions get made in one place and lost before they reach the next.
No amount of training fixes that. You can equip every person on your team with the skills to use AI well, but if the tools are fragmented, the value gets fragmented too. Enablement raises the ceiling for what your people can do. Friction lowers the floor of what the environment will allow.
Unrealized: Low friction + low enablement
When friction is low but enablement is weak, AI value goes unrealized. The tools and the environment are working, but your workforce hasn’t been equipped to take advantage of it.
This is where power users thrive in isolation. The engineers and PMs who seek out new capabilities on their own, who read the release notes and experiment on their lunch break, extract real value. But they’re the exception. The rest of the team uses AI for the basics—drafting, summarizing, searching—and stops there. Not because they can’t go further, but because nobody has shown them how or why.
Your organization invested in the tools and then underinvested in the people’s knowledge of the tools. Teams haven’t received the training to push past surface-level use, the context to understand what’s actually possible, or the time to explore. So capability accumulates in the hands of a few, and the broader workforce leaves value on the table every day.
Compounding: Low friction + high enablement
When friction is low and enablement is strong, your hard work pays off, and AI value compounds.
Your tools work together, and context travels with the work. Your teams know how to use what they have and keep finding new ways to get more out of it. Each improvement builds on the last. Value doesn’t stay locked with the people who go out of their way to use AI; it spreads across the team.
This is the state where AI stops being something your organization manages and starts being something that works for you.
Which state are you in?
Knowing which state your team is in changes everything about what you should prioritize next. Investing in enablement when friction is the real problem produces trained workers who are still working around a broken environment. Reducing friction when enablement is the gap produces a coherent stack that no one knows how to use.
The starting point is an honest read of where you actually sit.
The AI Value Quadrant Assessment maps your organization across both dimensions and surfaces what to prioritize first so you can start seeing value compound rather than stall.
