Beyond hard and soft skills

2026-09-09T00:00:00Z

Most job descriptions still split people into hard skills and soft skills. Most training budgets are spent along the same seam. Cold technicians and warm communicators, one column each, and a promotion process that quietly rewards the first.

It has been the default for decades and it is wrong in a way that costs you specific, nameable things.

Competence has three parts, not two. They are coupled, which means removing any one of them produces a person you have already worked with, and probably managed.

Why two was never enough

I was listening to Dave Snowden recently making the case for the trilectic. In a complex system, two opposing concepts are rarely enough to think with. A binary gives you a spectrum and a deadlock. You need a third term to break it and let the thing move.

He is not the first to find this. Lefebvre and Soja did the same for space when they refused the split between the physical and the mental and added the lived. The pattern repeats because two-part models are easy to hold and poor at describing anything that behaves like a system.

Human capability behaves like a system. So here is the third term, and the two you already know, described properly.

Bounded execution

This is what we have been calling hard skills. The disciplined application of known rules inside a clear boundary. Writing syntax. Running a financial model. Configuring a network. Reading a balance sheet.

This is where human energy converts to value with the least loss. The inputs are known, the constraints are known, and if you do the work well you get a predictable output.

It is the baseline. It is also nothing more than the baseline. Strength without coordination is someone who can lift a great deal and fall over regularly.

Interpersonal calibration

This is what we have been calling soft skills, usually with a faint apology in the voice. Empathy. Listening. Reading the room. Nice to have, HR will run a course.

Go back to the first law of thermodynamics. Energy cannot be created or destroyed, only transferred or transformed. Effort that does not become value becomes something else, and in a team that something is heat. Miscommunication. Ego. Ambiguous handoffs. The Slack argument on a Friday afternoon that everyone is still carrying on Monday.

Interpersonal calibration is the competence that keeps effort converting into value rather than transforming into friction. It is not a courtesy and it is not a personality trait. A team with it has working joints. A team without it has bone grinding on bone, and every movement costs more than it should.

Adaptive competence

Here is the part the binary cannot see.

Most training builds linear adaptation: tuning a skill to handle small shifts inside its own domain. The statistician learns a new package to run the same analysis. Useful, and not enough.

Complex environments demand something biologists call exaptation. A capability built for one purpose gets redeployed against an unrelated problem, because the underlying shape of the two problems turns out to be the same. An epidemiologist's contact-tracing model maps how risk cascades through a banking network. Nothing about the skill changed. The problem changed, and the skill fitted anyway.

Blockbuster is the counter-example everybody reaches for. Nine thousand stores' worth of bounded execution and not an ounce of adaptive competence when the ecosystem moved to streaming. Worth being precise about what failed there, because it is usually described as a strategy problem. It was not. The capability was present and the ability to redeploy it was not.

Nobody acquires a skill with foresight about where it will be exapted. The people who solve unfamiliar problems are carrying a store of capability built for something else entirely, and they are willing to test its fit when the pressure changes.

Three failure modes

Because the three are coupled, taking one away produces a recognisable person.

Missing adaptive competence gives you the stagnant specialist. Excellent execution, good with people, paralysed the moment the domain moves. This is the person promoted on the strength of a job whose parameters then changed entirely, and who is now failing in public at something nobody has acknowledged is a different job.

Missing calibration gives you the isolated executioner. Brilliant at the work, brilliant at the cross-domain leap, and generating so much heat that nobody can work alongside them long enough for any of it to scale. Every handoff costs more than it should. Teams reorganise around them rather than with them.

Missing bounded execution gives you the abstract visionary. Wonderful connections, charming in the client meeting, and no disciplined output at the end of it. Ideas that never harden into anything anyone can use.

You have worked with all three. You may have been one of them.

Why they are coupled

Read them together and the coupling is obvious.

Calibration lowers the cost of exaptation, because a cross-domain leap only lands if other people accept it. Bounded execution limits which exaptations are available at all, because you can only transfer a skill you hold at depth. And adaptive competence needs somewhere to happen: a team running at full velocity has no margin in which to attempt the leap, which is why the busiest teams are so often the most stagnant.

That last one matters more than it looks. Competence is not just a property of the person. It is a property of the person and the conditions they are working in.

Where this actually lives

The binary survives because it describes individuals, and individuals are what we recruit, appraise and promote. Two columns, one person, tick the boxes.

An organisation does not get the competence its people have. It gets the competence its structure allows them to use.

Adaptive competence needs slack, and slack is a structural decision. Calibration builds through repetition. People need to learn each other, and every reorganisation resets that to zero. Bounded execution is the only one of the three that survives a badly designed organisation intact, which is precisely why badly designed organisations end up valuing it above everything else. It is the part that still works when nothing else does.

So when a company tells me it has a capability problem, the interesting question is rarely whether the people are good enough. It is which of these three the organisation has quietly made impossible.

What to do with this

Stop assessing people on a two-column checklist. Competence is execution, calibration and transfer, and it is all three or it is none. Pick two and you have named your next failure mode.

Then ask the harder question. Look at the three, and work out which one your organisation is currently arranged to prevent.

P2O Consulting — Writing