Cynefin for HR: How to Stop Treating Complex People Problems Like Complicated Ones

60-Second Summary

Cynefin (Snowden; HBR 2007) sorts problems into: Clear (sense-categorise-respond), Complicated (sense-analyse-respond), Complex (probe-sense-respond), Chaotic (act-sense-respond), and Confused.
HR mostly operates in Complicated (compliance, payroll, systems) and Complex (culture, engagement, change, leadership development).
The classic error: treating Complex problems as Complicated ones — hiring a consultant to give the 'right answer' when only safe-to-fail experiments will produce learning.
Complex problems require probes: multiple small experiments, sensing what emerges, amplifying what works, dampening what doesn't.
Getting the domain right is the highest-leverage decision an HR leader makes on any initiative.

A CEO tells the CHRO: 'Fix the culture by Q3.' The CHRO hires a $400k consulting firm expecting a diagnosis and prescription. Nine months later there's a beautiful deck, a values refresh, and the culture is unchanged. The problem wasn't the firm — it was treating a complex problem as a complicated one.

The five domains

“The single most common failure of leadership is to apply the wrong sort of thinking to the wrong sort of problem.”
— Dave Snowden & Mary Boone, HBR (2007)

Domain	Nature	Approach	Example in HR
Clear	Cause and effect obvious; best practice exists	Sense → categorise → respond	Payroll processing, statutory compliance
Complicated	Cause and effect knowable with expertise	Sense → analyse → respond	Compensation benchmarking, HRIS implementation
Complex	Cause and effect only clear in retrospect	Probe → sense → respond	Culture, engagement, leadership development
Chaotic	No cause-effect discernible; novel practice	Act → sense → respond	Sudden crisis (leak, scandal, safety event)
Confused	Domain itself unclear	Break into parts and locate each	Any large 'transformation' before diagnosis

Where HR problems sit

A domain map of common HR work

1
Clear
Payroll runs, statutory filings, standard onboarding checklists, benefits enrollment mechanics.
2
Complicated
Comp philosophy design, HRIS choice, career ladder architecture, benefits plan selection. Expertise wins.
3
Complex
Culture change, engagement, leadership development, DEI outcomes, post-merger integration, remote redesign. Emergent — probes only.
4
Chaotic
Sudden scandal, viral employee story, unplanned mass layoff leak. Act first, stabilise into complexity as fast as possible.

Working each domain correctly

Complicated tools vs Complex tools

Complicated (works for HRIS, comp bands)

Hire expert, get analysis
Benchmark against best-in-class
Detailed rollout plan up front
Success = 'right answer' implemented
Big-bang deployment

Complex (works for culture, engagement)

Multiple small, safe-to-fail probes
Sense what actually shifts
Amplify what works, dampen what doesn't
Success = coherent direction that emerges
Portfolio of experiments, ongoing

How to run a probe (complex domain)

1
Design small and coherent
Small enough that failure is safe, large enough that a signal emerges. Trial a new 1:1 template with 30 managers, not 3000.
2
Instrument before launch
Know in advance what you'll measure — sentiment, behaviour proxies, qualitative themes.
3
Run several in parallel
Complex problems reward portfolios, not single bets. 3–5 probes attacking the same challenge from different angles.
4
Sense at defined intervals
Structured review at 30/60/90 days — what shifted, what surprised us, what emerged that we didn't predict.
5
Amplify or dampen — don't 'evaluate for rollout'
The complex-domain move isn't 'pilot then rollout'. Amplify the ones creating good emergence, dampen the rest, keep probing.

Common category errors

Treating culture as Complicated — hiring a consultant to deliver the 'right' values. Culture is emergent; you can only probe.
Treating engagement as Clear — buy a platform, roll it out, done. Engagement is Complex.
Treating a crisis as Complicated — waiting for a full analysis while the situation deteriorates.
Treating payroll as Complex — inventing bespoke processes when a standard best-practice solution exists.
Treating post-merger integration as a project plan — the human integration is deeply Complex.

The tell

If you demand 'the right answer' before starting, you think you're in the Complicated domain. If the problem is Complex, the right answer doesn't exist yet — it will emerge from probes.

FAQ

Frequently asked questions

How do I convince the CEO complex problems don't have neat answers?

Show historical patterns — every 'culture reset' run as a Complicated project, and the win rate. Then propose a probe portfolio with defined 90-day sensing points.

Isn't 'probe-sense-respond' just Agile?

Related, not identical. Agile applies iteration to complicated product work; Cynefin insists Complex problems are ontologically different.

Where does data-driven HR fit?

Beautifully in Clear and Complicated. Cautiously in Complex — data helps sense what's emerging, but past patterns don't guarantee future ones.

Takeaways

Getting the domain right is the highest-leverage decision on any HR initiative.
Culture, engagement, leadership development, and change are Complex — probe, don't prescribe.
Complicated tools applied to Complex problems produce beautiful decks and no change.
Run portfolios of small probes, sense what emerges, amplify what works.