Free diagnostic · no sign-up

How does your system behave under stress?

Not how much stress it is under. Six archetypes describe how an industrial operating system responds when something changes — whether it absorbs the disturbance, amplifies it, cycles on it, or stops responding at all. Each implies a different decision architecture.

Taxonomy from the Industrial Genome system-identification model
What this is, and what it is not

Genome classifies a MEASURED fingerprint derived from external signal data. The self-assessment described below classifies SELF-REPORTED answers using the same criteria and priority order. They share a vocabulary, not a method. A self-assessment result is a recognition aid; it is not a fingerprint and should not be reported as one.

The six dimensions

What gets measured.

Genome infers these for public industrial companies from external signals. Latency is the one the rest of this site is about; the other five explain why a given latency costs what it costs.

The six archetypes

Six ways a system answers a disturbance.

Saturated

Throughput ceiling detected. Increasing inputs stops yielding proportional output.

SignatureHigh gain with high damping and high volatility transmission — running at capacity, amplifying inputs but still absorbing shocks.
Where is the constraint, and what is the debottlenecking investment needed?

The decision that matters is capital, and it is usually waiting on an evidence standard nobody has defined. Pre-authorize reversible experiments before the full capital case.

Fragile

System amplifies shocks. Small external perturbations cause disproportionate operational disruption.

SignatureTop-quartile volatility transmission with top-quartile gain. These systems propagate stress rather than contain it.
What is the primary amplification path, and what buffering or diversification removes it?

Latency is most expensive here, because consequence compounds fastest. This is the archetype where permission in advance pays for itself.

Oscillatory

Demand or inventory cycling. Boom-bust pattern in evidence. Bullwhip effect likely.

SignatureHigh oscillation frequency with low damping, or sustained moderate volatility transmission.
What is driving the cycle, and can it be dampened with demand visibility or vendor-managed inventory?

The cycle is usually manufactured by the decision path itself: each function reacts to the last function's correction. Shorten the loop before adding forecasting.

Constrained

System is hitting a structural limit — capacity, supplier, or capital constraint binding.

SignaturePositive but moderate gain with low damping. Output grows sub-linearly; something is binding.
What is the binding constraint, and what is the cost of removing it?

Everyone can name the constraint and no one owns removing it. This is a decision-rights problem wearing an engineering costume.

Resilient

Well-buffered. System absorbs external signals without proportional disruption.

SignatureHigh damping with low volatility transmission and gain below one.
Where is the resilience coming from, and can it be a competitive advantage?

Worth knowing whether the buffer is architecture or inventory. Inventory-bought resilience is latency you are paying for in working capital.

Decoupled

Observable signals have low explanatory power. Internal data or a different signal mix is needed.

SignatureVery low gain, or low confidence across dimensions — output does not respond to standard external inputs.
What internal metrics would reveal the actual constraint?

Not a diagnosis so much as an admission that the instrument cannot see. Usually means the binding constraint is internal and undocumented.

The blunter cut

A second, blunter cut, from Michael Carroll's Bill Comes Due framework.

Does the system intervene at the commitment point, or does it record what happened after the fact?

The architecture carries the work

The system intervenes at the commitment point. People supervise it rather than compensate for it.

Latency at or below the industry median, with damping above 0.5.

People carry the work

The system records what happened. Humans are the integration layer, and their capacity is the ceiling.

Latency above the industry median together with cycling, or latency above 1.5x the median on its own.

Self-assessment

Which one is yours?

Five questions, answered about one decision family rather than the whole company. Same criteria and priority order Genome uses, applied to what you report instead of to measured signal data.

days

Does adding more input — more people, more hours, more capital — stop producing proportionally more output?

Does a small external disruption produce a disproportionately large operational one?

After a disruption, does the system settle quickly, or keep ringing for weeks?

Is there a recurring boom-bust or inventory cycle — building up, correcting, building up again?

Can performance be explained from external conditions, or does it seem disconnected from them?

Closest archetype

Benchmark

Median response lag, by industry.

The medians the closure classification is measured against. Compare a real signal-to-stabilized-response time before concluding anything about your own speed.

IndustryMedian days
semiconductors20
metals22
food beverage24
industrial automation25
hvac26
construction26
diversified industrial27
auto28
building materials28
general28
heavy equipment30
chemicals30
oil gas30
agricultural equipment32
aerospace defense35
electric utilities35
Where this goes

The archetype tells you which decision to redesign first.

A saturated system and a fragile one both feel slow from the inside, and they need opposite interventions. The masterclass takes one decision family and rebuilds its path; the archetype is how you choose which one.