Don’t Start with AI. Start with Friction.

Every organization is under pressure to adopt AI. The tools are proliferating, the case studies are compelling, and the fear of falling behind is real. But in the race to implement, most teams are asking the wrong first question.

They ask: “where can we use AI”?  when they should be asking: “where are we stuck”?  The distinction matters more than it might seem.

AI is an accelerant. Like any accelerant, it amplifies what is already there. Apply it to a clear, well-structured workflow and it can meaningfully increase speed, reduce repetitive effort, and improve consistency. Apply it to a chaotic or unstable process and it will amplify the confusion, the rework, and the inefficiency, just faster and at greater scale. The technology does not fix broken systems. It exposes them.

Customer education teams feel this acutely. The work is inherently complex: translating deep product knowledge into training that sticks, keeping pace with release cycles that rarely slow down, and depending heavily on subject matter experts (SMEs) whose calendars are never as available as the project plan assumes. These are not small inefficiencies. They are structural friction points, and they are exactly where intentional AI use can make a meaningful difference.

“The technology does not fix broken systems. It exposes them.”

This is the foundational insight behind SmartScAIle™, a framework I developed for accelerating customer education content development with AI. Rather than starting with a tool and looking for a problem to solve, SmartScAIle starts with a structured diagnostic that surfaces where workflows break down across five stages: clarifying inputs, accelerating drafting, standardizing and reusing content, streamlining reviews, and sustaining updates over time.

The diagnostic is deceptively simple. Teams rate themselves across each stage, then rank where their greatest constraint lives. What emerges is often surprising. A team convinced its bottleneck is slow content drafting discovers the real problem is upstream: unclear objectives, undefined scope, or a product release that outpaced the content plan before a single word was written. Another team investing heavily in AI writing tools finds that drafts are fine. The real constraint is SME availability. Knowledge lives with a handful of people whose time is scarce, which means content sits waiting, gets written with incomplete information, or requires multiple revision cycles to get right.

This matters because AI tools are often purchased to solve the symptom, not the constraint. And when the constraint is not addressed, the symptom returns, now with a subscription attached.

“This matters because AI tools are often purchased to solve the symptom, not the constraint.”

According to Asana’s Anatomy of Work Index, knowledge workers spend roughly 60% of their time on “work about work”, searching for information, coordinating tasks, and chasing approvals, rather than doing the work itself. In customer education, that often looks like tracking down the right SME, reconciling conflicting product documentation, or rebuilding content that should have been reusable but was not designed that way. That is not a tools problem. It is a workflow problem. And it is exactly the kind of structural friction that, when addressed intentionally, creates the conditions for AI to deliver its real value.

The opportunity is substantial. McKinsey’s research on the economic potential of generative AI found that the technology could automate 60 to 70 percent of time currently spent on routine tasks, with its greatest impact concentrated in areas like content creation, summarization, and editing. For customer education teams managing growing content libraries against accelerating product roadmaps, that is a meaningful ceiling to reach for. But reaching it requires something the tools themselves cannot provide: a clear view of where the work actually breaks down.

“The teams that will get the most from AI in the years ahead will not necessarily be the ones who adopted it earliest.”

The SmartScAIle™ approach does not resist AI adoption. It sequences it. Find your constraint first. Ask whether the workflow in that stage is structured enough that someone new could follow it. Ask where decisions stall, where handoffs create confusion, and where work moves forward only to be pushed back for revision. Then, and only then, ask how AI could help.

That sequencing is what makes AI implementation sustainable rather than theatrical. It shifts the question from ‘Are we using AI?’ to ‘Is AI solving the right problem?’ For customer education teams navigating growing demand, shrinking timelines, and constant pressure to do more with less, that distinction is the difference between a tool that transforms the work and one that simply adds to it.

The teams that will get the most from AI in the years ahead will not necessarily be the ones who adopted it earliest. They will be the ones who were intentional about where they applied it, who took the time to diagnose before they deployed, and who understood that friction, not technology, is usually the real constraint.

“The point isn’t to use AI for its own sake; it’s to build repeatable ways of working that make our learning ecosystem easier to scale.”

And to answer the obvious question: yes, we are using the SmartScAIle™ model at Charles River, starting within Global Educational Services, to surface where our content workflows stall and where AI can responsibly remove drag. The point isn’t to use AI for its own sake; it’s to build repeatable ways of working that make our learning ecosystem easier to scale.

By diagnosing the main constraint before implementing AI, whether we are developing and maintaining training materials or delivering them through CRD Academy, we ensure that AI technology is deployed where it provides genuine value. This disciplined approach helps us accelerate progress, reduce unnecessary revisions, and keep our training content in step with current versions of our software.

Sources:

Asana Anatomy of Work Global Index 2023. asana.com

McKinsey Global Institute. The Economic Potential of Generative AI: The Next Productivity Frontier. June 2023. mckinsey.com