Before applying your usual management instincts, ask: is this work replicating known value or discovering new value?
The answer changes everything about how you should manage it.
The two modes
Factory mode (replication): You know what good looks like. The goal is consistency, efficiency, reducing variance. Measure output. Set targets. Optimise the process.
Examples: onboarding, billing, support, manufacturing, fulfilment.
Lab mode (discovery): You don’t know what good looks like yet. The goal is learning speed — finding what works through experimentation. Measure what you learned, not what you shipped. Tolerate failure.
Examples: new product development, entering new markets, solving novel problems, early-stage ventures.
The trap
Most leaders default to factory mode. It’s what business training teaches: deadlines, utilisation, efficiency metrics, accountability for results.
These instincts actively harm discovery work. Arbitrary deadlines encourage shipping before you’ve learned enough. Efficiency metrics punish the experimentation that generates insight. Accountability for results creates risk-aversion when you need risk-taking.
The result: teams optimise for looking productive instead of actually learning. They ship something by the deadline, but they haven’t discovered anything. The work looks complete; the value isn’t there.
The check
Before starting work — or when something feels stuck — run through three questions:
What kind of work is this? Do we know what success looks like, or are we trying to find out?
Am I managing it with the matching mode? Factory work needs process discipline. Lab work needs learning discipline. Using factory tools on lab problems breaks them.
What am I actually measuring? Factory: output, efficiency, variance reduction. Lab: experiments run, hypotheses tested, insights generated.
If you’re running lab work like a factory, you’ll get compliance without discovery. If you’re running factory work like a lab, you’ll get chaos without consistency.
Match the mode to the work.
Related: Hidden Bottleneck · The Last Twenty Percent