Simple rules beat complex plans

In a complex environment, a short list of clear decision rules outperforms a detailed plan, because plans assume the future and rules respond to it. A plan tells you what to do in the situation you predicted; a rule tells you what to do in the situation you are actually in. The practical move for a founder is to replace the annual strategy document nobody reads with five to ten if-then rules the whole firm can apply without asking.

Why do complex plans fail in complex systems?

A plan is a prediction wearing a suit. It encodes assumptions — about the market, the team, response rates, the competitor's next move — and a business is a system with feedback loops, delays and adaptive people on every side of every transaction. As I set out in the systems-thinking guide for founders, you cannot predict such a system in detail; you can only understand its structure and act on it. So the 40-page plan is typically wrong by week three, and then something worse happens: because it cost so much to produce, people keep executing it anyway. The plan stops being a map and becomes a commitment device for ignoring feedback.

Rules fail differently. A rule is small, so it is cheap to test, cheap to break, and cheap to replace. When a rule turns out to be wrong, you amend one line, not one strategy.

What makes a rule actually useful?

Three properties, and most "company values" have none of them.

• It triggers on something observable. "Below 3% positive replies" is observable. "When quality slips" is not.
• It names one action. Fix the campaign. Raise prices. Call within five minutes. Not "review and consider options".
• It removes a decision from a queue — usually the founder's. If a rule still requires judgement from the busiest person in the firm each time it fires, it is a suggestion, not a rule.

The mechanism is worth stating plainly, because it explains why rules compound. When a decision recurs, you write a rule for it. When the rule exists, the decision executes at the point of work, without escalation. When decisions stop escalating, the founder's queue shortens and the system's delays shrink. When delays shrink, feedback arrives faster, so the next rule is written on better information. Each rule you write improves the conditions for writing the next one.

Which rules do we actually run a pipeline on?

These are working rules from our own outbound and sales system, not illustrations:

None of these is clever. That is the point. Each one converts a recurring judgement call into a reflex, which means it happens on the bad weeks too — and the bad weeks are when judgement is worst.

Don't simple rules get gamed?

Any rule can be, which is why the rules above trigger on outcomes rather than proxies wherever possible — positive replies rather than opens, win rate rather than proposals sent. When a measure becomes a target it stops measuring; I have covered that failure mode in detail in Goodhart in the wild. The defence is the same in both cases: few rules, honest triggers, and a periodic review where you ask which rule most recently produced a stupid outcome. Rules are a stock; review is the flow that keeps them from decaying.

The other objection is that simple rules sound fragile — surely a sophisticated plan handles shocks better? In practice the opposite holds. A firm running on a detailed plan absorbs a shock as damage, because the plan assumed the shock away. A firm running on rules absorbs it as information: the triggers fire, the responses execute, and the rule set gets one line better. That is a large part of how you build a pipeline that benefits from shocks.

Where should a founder start?

Not with a blank page. Start with your last ten interruptions — the questions people brought you, the deals you personally rescued, the leads that sat for three days. Each one is a rule waiting to be written. If the same rescue keeps recurring, that is also a signal the underlying system needs rebuilding rather than another patch; a CRM full of ignored leads is the classic example, and it responds to a reactivation rule ("no reply in 90 days, enter the reactivation sequence") faster than to any strategy offsite.

Write five rules. Run them for a quarter. Delete the ones nobody used. What remains is your actual strategy — the one your firm executes when you are not in the room.

---

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