You Can't Fix a Complex Problem with a Complicated System

The handshake happens near the end of the launch or training day. The room is warm from a full afternoon of slides and worked examples. Someone from the leadership team stands up, thanks everyone, and the sellers nod along.

The tracking system is live by then. Fields to fill, a planner attached to each opportunity, a weekly cadence agreed. It feels like commitment. You can see it in people's faces. They understood the case, and they bought it.

Then you come back a few months later.

Nothing has collapsed. There is no revolt, no formal complaint, no meeting where anyone declares the thing a failure. What you find instead is quieter. The planner is still there. Sellers open it on a Thursday afternoon and complete it before Friday, the way they clear a CRM field they have decided not to think about.

The energy from launch day is gone. The tool survived. The thinking behind it did not.

I have watched this scene repeat across industries with almost nothing else in common. Medical devices. Enterprise software. Industrial supply. Financial services. The setting changes. The pattern does not.

The Explanations We Reach For

When adoption fades, the search for a reason starts quickly, and the usual suspects are always close to hand.

Seller resistance. Managers who did not hold the line. A culture with no accountability. The wrong technology, badly configured. Training that was too thin to stick.

None of these are wrong, exactly. Each one is real, and each one shows up in the post-mortem for a reason. The problem is that they describe what you can see on the surface. They name the visible breakdown and treat it as the cause.

A seller ticking a box without thought is a symptom. Weak enforcement is a symptom. Fix a symptom and you get relief for a while. Then the same drift returns wearing slightly different clothes.

Something more fundamental is being misread. It sits underneath the behaviour, and it rarely gets named in the room where the fix is decided.

The Single Stakeholder Trap

Here is one place the misreading becomes visible.

Think of the seller who has built a career on one relationship. In medical devices it is the surgeon. In enterprise technology it is the CIO or the CISO. In large commercial deals it is the CFO. One dominant contact, cultivated over years, who opens doors and signs off spend.

That relationship works. It has worked well enough to build a livelihood on, to hit targets on, to earn a reputation on. The seller is not clinging to a bad habit. They are trusting what has repaid their trust for a decade.

Then a programme arrives and asks them to think more broadly. Map the wider buying group. Question the strength of the single relationship. Test whether their preferred stakeholder actually holds the power they assume.

To the person receiving it, this does not feel like an upgrade. It feels like a challenge to what they know to be true. Their instinct pushes back, and from the outside that looks like resistance.

It is closer to rationality. The seller has good evidence that their approach works, and you are asking them to bet against it. That kind of doubt is stubborn precisely because it is not really about process. The mandate can't touch identity. A directive can change what someone does on a Thursday afternoon. It cannot change what they believe about their own way of winning. You can require the form to be filled. You cannot require someone to stop believing the thing their own success taught them.

Complicated Is Not Complex

This is the distinction I keep returning to, because most adoption failures live inside it.

A complicated thing has many parts, but the parts behave predictably. You can write a manual for it. A jet engine is complicated. A tax return is complicated. Build the right system, apply it consistently, and you get the result you designed for. The difficulty is real, and it is solvable with rigour.

A complex thing is different in kind. Its parts interact in ways you cannot fully predict, and they adapt to the conditions you create. Change one element and the system responds, sometimes in a direction you did not intend.

Human organisations are complex. Sales cultures are complex. A single seller, shaped by ten years of habit and belief and one industry's particular power structure, is deeply complex.

The programmes we build to change them are complicated. Fields, cadences, checklists, review rhythms. Well made, internally coherent, and complicated. We install them inside complex environments and then express surprise when they do not hold.

The Book and the Child

There is a version of this that anyone with children will recognise.

A good book on parenting is genuinely useful. It was written by people who thought hard about the problem, drew on real observation, and offered advice that holds up. The trouble is straightforward. The child has not read it.

The parent who applies the book to the letter runs into developmental stages that arrive off schedule, a friendship group pulling in its own direction, a bad week at school, and a personality that was never going to match the template. None of this means the book is wrong.

It means the book is a complicated answer to a complex situation.

Sales adoption works the same way. The mandate is the book. The seller is the child. And the world keeps happening around both of them, refusing to sit still long enough for the manual to apply cleanly.

What This Actually Means

Before the system, before the fields and the cadence, there is a prior question about the person.

Who are they. What do they believe about their customers and about their own way of winning.

And then the question that most programmes never ask: what would adopting this cost them in their sense of who they are at work.

That question is what separates an adoption problem from a process problem. It is also the one that gets skipped, because the process is the part you can design, schedule, and measure. The person is harder to see and impossible to standardise, so the person gets left out.

That is why the handshake enthusiasm evaporates. The launch spoke to belief for an afternoon. The programme that followed spoke only to behaviour. Belief quietly left the room while the activity carried on, because compliance is cheap and conviction is not.

Where This Leaves Us

If you have built one of these programmes, you almost certainly built it in good faith. You saw the problem clearly, designed a rigorous response, and felt the energy in the room on launch day. The drift that followed was not evidence of poor thinking. It was evidence of a harder problem than the one you were given permission to solve.

The organisations getting closer to making adoption stick are not the ones with the most detailed tracking. Their dashboards are not what sets them apart.

They are the ones who understood the complexity first. They looked at the seller as an adaptive person with a working identity, and they built their intervention around that reality rather than over the top of it.

That is not a programme you can buy and install. It is a different kind of leadership, slower and less certain, more interested in what a person believes than in what a form records.

We keep building complicated systems for a complex world, and then we return months later to the quiet drift. The system was sound. We simply never asked the harder question about the person standing behind it.

A Transparency Note

I should be honest here, because the risk of a post like this is that it describes a problem I have conveniently solved.

I have spent years working on exactly this challenge. Peer cohorts. A thinking discipline that sits underneath the deal. An AI that coaches the seller's thinking rather than records their activity. I am aware that any of them, applied as a mandate from above, become precisely what this post describes. The planner becomes a CRM field. The cohort becomes a compliance session. The AI becomes something sellers open on a Thursday afternoon.

The difference — and I say this having watched it go both ways — is not in the technology. It is in where the work starts. The cohorts that produce something durable begin with a peer conversation about what sellers actually believe, not a directive about what they should do. The AI that gets used honestly starts with the seller asking what they know versus what they are assuming. That is a different starting point than a process. It is the prior question this post keeps returning to.

I do not have a programme that resolves the complexity problem. No one does. What I have is a starting point that at least faces the right direction.