← All essays

Organizations

The Photocopier Problem

Ken Riley  ·  March 2026  ·  4 min read

There’s a concept in organizational behavior so common it barely gets named anymore. I call it the photocopier problem — and once you see it, you cannot unsee it.

Here’s how it works. An organization develops a process to solve a specific problem. The process works. People learn it, refine it, get good at it. Over time, that process becomes institutionalized — it gets documented, trained, measured, and eventually protected. The people who operate it build their identity around their mastery of it.

Then the original problem changes. Or disappears entirely. Or gets replaced by a different, more important problem.

The process keeps running.

Think about how a photocopier degrades a document. The first copy is nearly identical to the original. The second copy is a copy of that copy. By the tenth generation, you have something that looks like the original document — has the same general shape, the same approximate content — but every line is slightly softer, every edge slightly blurred. The drift is so gradual that no single copy seems obviously wrong. The problem only becomes visible when you hold the tenth copy next to the original.

Most organizational processes work exactly this way. Nobody lies. Nobody cheats. Nobody makes a deliberate decision to preserve something useless. What happens is subtler and more human: people become invested in the thing they know how to do. They optimize it. They defend it. They hire people who reinforce it. And slowly, without anyone noticing, the process becomes the point — rather than the outcome the process was designed to create.

I have seen this in health systems where clinical workflows designed for paper records were still shaping the architecture of digital systems twenty years later. I have seen it in financial institutions where risk processes built for a regulatory environment that no longer exists consumed enormous resources and generated reports nobody read. I have seen it in technology organizations where entire teams were organized around products that had been strategically deprioritized two leadership cycles ago.

In every case, the people inside the system were competent, committed, and working hard. The problem wasn’t individual performance. The problem was that the system itself had drifted so far from its original purpose that nobody inside it could see the distance anymore.

This is why proximity is so dangerous for organizational thinking. The closer you are to something, the harder it is to see what it actually is versus what it was designed to be. The people most invested in a process are precisely the people least equipped to question it — not because they’re incapable, but because their identity, their expertise, and often their job security are built around it continuing to exist.

The photocopier problem doesn’t get solved by working harder. It doesn’t get solved by better process documentation, more rigorous performance metrics, or a new leadership framework. It gets solved by one thing: someone willing to walk into the room and ask the question nobody has asked.

What problem was this designed to solve? Does that problem still exist? Is this still the best way to solve it?

These are not complicated questions. They are, however, uncomfortable ones — because the honest answer often requires dismantling something that a lot of people have built their professional identity around. That’s not a technical challenge. That’s a human one. And it requires a particular kind of presence in the room: someone with no stake in the answer, no political position to protect, and the ability to hold the discomfort of the question long enough for the real answer to emerge.

The most useful thing an outside advisor can do for an organization is not bring answers. It’s bring the question that’s been too dangerous to ask from the inside.

The photocopier is still running. Someone needs to ask what it’s copying.

Ken Riley is a technology executive, strategic advisor, and board member with 20+ years of experience leading organizations through complexity. He works with leaders and organizations who need a trusted thinking partner — someone who will ask the question nobody else is asking.

← All essays

Healthcare IT & AI

What Your Patients Are Already Telling You — And Why No One Is Listening

Ken Riley  ·  2026  ·  5 min read  ·  Research paper below

Every time a patient sends a message through your portal, they’re telling you something about how they think about their health. Not just what they’re asking — how they’re asking it. The words they choose. The questions they lead with. Whether they’re assertive or deferential, proactive or reactive, focused on information or focused on reassurance.

That signal is psychographic data. And most health systems are generating it by the millions of messages — and doing nothing with it.

For the past two decades, healthcare has operated on a one-size-fits-all communication model. The same discharge instructions. The same appointment reminders. The same health education materials — delivered to every patient regardless of how they actually process information and make decisions about their health.

We know this doesn’t work. Chronic disease management programs that achieve 10% improvement through personalized digital engagement are celebrated as successes — which tells you something about how low the baseline is. The gap between what patients need to hear and what health systems say to them is enormous. And it’s not a content problem. It’s a segmentation problem.

The five psychographic segments that matter in healthcare — Self-Achievers, Balance Seekers, Priority Jugglers, Direction Takers, and Willful Endurers — each require fundamentally different communication strategies for the same desired outcome. Tell a Self-Achiever to “achieve their health goals” and they respond. Tell a Willful Endurer the same thing and you’ve lost them. The message that activates one segment actively repels another.

The challenge has always been scale. Psychographic segmentation works — the research is clear on this, including a model that achieves 91% accuracy in segment identification. But traditional segmentation requires surveys. And getting patients to complete surveys at meaningful scale is, to put it generously, difficult.

This is where natural language processing changes everything.

Patient portal messages are already written. They exist by the millions in systems that healthcare organizations already own. And they contain — embedded in word choice, sentence structure, tone, and query type — exactly the psychographic signals that segmentation models need to classify a patient’s motivational profile.

You don’t need to ask patients who they are. They’re already showing you. The patient who messages “I’ve been researching my medication and I have some questions about the clinical evidence” is showing you a very different psychographic profile than the patient who messages “my doctor told me to take this, can you remind me what it does.” Both need information. Neither needs the same information delivered the same way.

My doctoral research explored exactly this intersection — using NLP to segment patient populations through the messages they’re already sending, enabling healthcare systems to activate psychographic personalization at scale without requiring a single additional survey or data collection step.

The implications extend beyond patient communication. The same framework applies to any organization operating in a regulated environment where understanding the motivational drivers of the people you serve — patients, customers, members — is the difference between engagement that moves behavior and messaging that disappears into noise.

The data already exists in your systems. The question is whether your organization has built the capability to listen to what it’s saying.

Ken Riley has spent 20+ years at the intersection of healthcare technology, patient engagement, and organizational strategy. His doctoral research focused on psychographic segmentation through NLP of patient portal messaging. He is currently VP of Customer Success at Psympl, an enterprise Psychographic AI platform for financial services.