The system archetypes founders keep living

System archetypes are recurring loop structures that produce the same business problems in firm after firm, regardless of industry or effort. Learn perhaps five of them and you can diagnose most stalled growth stories in an afternoon, because the pattern — not the people — is generating the behaviour. The practical payoff is speed: instead of debating personalities, you name the archetype, find the loop, and change the structure.

What is a system archetype?

A named, repeating arrangement of feedback loops. Donella Meadows calls them systems traps in Thinking in Systems; Peter Senge catalogued them as archetypes in The Fifth Discipline. The claim behind both books is the same and worth taking seriously: when many unrelated organisations produce identical dysfunctions, the cause is structural, not personal. As I argued in the systems-thinking guide for founders, structure produces behaviour — so recognising the structure is most of the diagnosis. What follows are the four archetypes I meet most often inside UK B2B service firms of 5–50 staff.

Limits to growth: why did we stall at this size?

A reinforcing loop drives growth — referrals bring clients, clients bring referrals — until a balancing loop it was quietly feeding switches on: delivery capacity, the founder's calendar, the size of the network being farmed. Growth flattens, and the instinctive response is to push the old engine harder. It cannot respond; the constraint is elsewhere. Goldratt's The Goal makes the operational version of this point — output is set by the bottleneck, so effort anywhere else is cosmetic.

The mechanism is compact. When the growth loop runs, it consumes some finite capacity; when that capacity approaches its limit, each unit of push yields less; when yield falls, firms typically push harder on the same lever, which consumes the capacity faster still. The exit is never more push — it is finding and expanding the constraint. A referral-dependent firm has usually hit the limit of its network, not the limit of its market; the arithmetic of that ceiling is laid out in the maths of referral-only pipelines.

Eroding goals: when did "below 3% means fix it" become "3% is fine"?

Every target has two ways to close the gap with reality: improve performance, or lower the target. Lowering is quicker and quieter, so under sustained pressure standards drift — response times stretch, the definition of a qualified lead loosens, "we reply to every lead within five minutes" becomes "same day" becomes "when someone gets to it". Each individual adjustment is defensible. The compound effect is a firm whose standards are set by its worst recent quarter.

The defence is mechanical, not motivational: anchor standards to something outside this month's comfort. Written thresholds with pre-committed responses — below 3% positive replies, fix the campaign; win rate above 60%, raise prices 15% — cannot drift silently, because the drift would require editing a document rather than merely relaxing a habit.

Shifting the burden: why does the quick fix keep winning?

The rescue relieves the symptom, the relief removes the urgency, and the capability for the real fix atrophies — the founder-as-hero loop being the canonical small-firm case. I have unpacked this one fully in why fixes fail, so here I will only place it in the family: it is the archetype that most often hides inside the others. Firms stuck at a growth limit are usually also shifting the burden — the founder's rescues are the patch that makes the missing pipeline system survivable one more quarter.

Success to the successful: why does one channel get all the investment?

Whatever performed last quarter gets more budget and attention this quarter, which makes it perform better relative to the starved alternatives, which justifies the next allocation. The reinforcing loop feels like discipline — back the winner — but it manufactures concentration: one channel, one rainmaker, one anchor client. The winner did not win on a level field; it won on compounding allocation. The result is a fragile pipeline with a single point of failure, and the case for deliberately funding parallel engines — even when one is "obviously" better — is the case I make in a pipeline that benefits from shocks.

How do you actually use the archetypes?

A three-step habit, twenty minutes, done whenever a problem has survived two previous fixes.

1. Plot the behaviour over time. Stalls, drifts, oscillations and deepening dependencies each have a characteristic curve. The shape suggests the archetype.
2. Name the loops. What is reinforcing, what is balancing, and where the delays sit. If you cannot name the loop, you have a symptom, not a diagnosis.
3. Change structure, not effort. Expand the constraint, write the threshold down, ring-fence the fundamental fix, fund the second channel. If the intervention amounts to "try harder", it is not an intervention.

The recurring lesson of all four archetypes is that firms do not fail at growth for lack of effort; they run structures whose behaviour is stalling, drifting or concentrating, and then exhaust themselves fighting the output instead of the structure. Some structures, though, you want on your side — the reinforcing loops that compound in your favour. That is the subject of the next piece, flywheels vs funnels.

---

Keep the thinking coming: The Format — a fortnightly letter on business and personal systems. Subscribe by email.