What Startups Teach Enterprises About Speed

McKinsey's global survey work on organisational decision-making found that only around a fifth of respondents believed their organisations excelled at decision-making, and that respondents at organisations described as decision-making winners were roughly twice as likely to report that their significant decisions delivered superior financial returns. These are self-reported survey results rather than measured outcomes, and self-assessment is a weak instrument. But the direction matches what I have watched happen in operating businesses.

The same body of work makes a point that deserves more attention than it gets: speed and quality are not necessarily a trade-off. Organisations that decide quickly are frequently the same ones that decide well, because the practices that produce speed, clear ownership and a defined evidence threshold, also produce clarity.

The cost of latency shows up in seven places, and none of them appear on a dashboard.

• Lost market windows, where a competitor establishes the default before you launch.
• Repeated analysis, where a question is re-examined because the last answer expired before it was used.
• Internal workarounds, where teams build shadow processes to avoid the official one.
• Diffused accountability, because a decision made by a committee is a decision owned by nobody.
• Employee frustration and attrition among exactly the people who could have moved fastest.
• Delayed customer learning, where the feedback that would have improved the proposition arrives after the budget is committed.
• Compounding execution cost, because late decisions are implemented under time pressure and therefore implemented badly.

Building a vehicle subscription business across the UAE, Saudi Arabia, Singapore and Thailand taught me more about decision speed than any framework I had read.

Limited capital forces prioritisation in a way that no strategy process replicates. When there is enough money for one of three initiatives, the debate resolves quickly because the alternatives are visibly mutually exclusive. Large organisations lose this discipline precisely because they can afford all three, and then wonder why none of them landed.

Market feedback replaces abstract debate. In a small business, a disagreement about pricing can be settled in ten days with real customers. In a large one, the same disagreement can absorb three months of analysis and still be settled by seniority.

Operational reality tests every proposition. In an asset-heavy model, a marketing promise that the operation cannot deliver is exposed immediately, at the vehicle handover. This is a gift. Most enterprises have far longer gaps between promise and delivery, which allows flawed propositions to survive far longer.

Cash flow makes delayed learning expensive. Every month of ambiguity has a price you can read on a bank statement. It focuses attention in a way that a variance report never does.

International expansion exposes assumptions faster than anything else. Entering a new market compresses years of assumption-testing into a quarter, because everything that was implicit in the home market becomes an explicit question.

And the most useful distinction of all: reversible versus irreversible. We could change a price, a bundle or a channel in days and reverse it if it failed. We could not easily reverse a fleet purchase, a market entry or a senior hire. Applying the same governance weight to both categories is the most common way that organisations become simultaneously slow and reckless: slow on the small things, insufficiently careful on the large ones.

Treat reversible decisions as experiments and irreversible ones as commitments. Most organisations do the opposite without realising it.

• Clear ownership. One named person per decision, published, with authority that matches the accountability.
• Smaller cross-functional teams that contain every function needed to move, so progress does not depend on another team's queue.
• Time-boxed analysis. Analysis expands to fill the time available, so the time available must be set in advance.
• Direct customer feedback for the people making the decision, not filtered through three layers of summary.
• Shorter operating cadences. Weekly beats monthly for anything that is still being learned.
• Explicit success measures agreed before launch, which also makes stopping politically survivable.
• Faster escalation paths, with a standing rule that an unresolved decision escalates automatically after a defined period rather than waiting for someone to raise it.

Startup admiration is often uncritical, and some of what makes small companies fast is simply immaturity that has not yet become expensive.

• Weak governance, which is survivable at ten million in revenue and catastrophic at a billion.
• Founder dependency, where every meaningful decision routes through one person. This is a ceiling, not a strength.
• Constant priority changes described as agility. Reprioritising weekly is not speed, it is churn.
• Poor documentation, which converts institutional knowledge into personal knowledge and loses it at the first departure.
• Inadequate risk assessment in domains where the downside is regulatory, safety-related or systemic.
• Hero culture, which rewards rescue rather than prevention and quietly punishes the people who make things boring.
• Unsustainable workloads, which produce a burst of velocity followed by an attrition bill.

Acceleration in a large organisation is an architecture problem rather than an attitude problem. Four levers do most of the work.

Decision rights. Write down who decides what, at which value threshold, and publish it. A surprising share of organisational slowness is simply uncertainty about authority, and it is resolved by a document rather than by a transformation.

Risk-based governance. Tier decisions by reversibility and exposure. Full committee scrutiny for the irreversible and material. A single approver with post-hoc review for the reversible and small. The current practice in many companies, one process for everything, guarantees the wrong allocation of scrutiny.

Escalation thresholds. Define in advance what triggers escalation, so that escalation is a designed pathway rather than a political act.

Delegated authority close to the customer. The person with the most information about a customer situation is usually the furthest from the authority to act on it. Closing that gap, even modestly, is one of the highest-return interventions available to a commercial leader.

Speed amplifies whatever the system already is. In an organisation with poor information, misaligned incentives or unclear ownership, acceleration produces more mistakes faster, and some of those mistakes are not recoverable.

In regulated industries the constraint is often legitimate rather than bureaucratic. Controls that look like friction from a commercial seat frequently exist because a regulator, a court or a previous incident put them there. The right response is to make the control efficient, not to route around it.

There is also a strategic consistency risk. A portfolio of fast local decisions can add up to an incoherent enterprise position, particularly in pricing, brand and partnerships. Delegation needs guardrails that describe the boundaries of acceptable variation.

The distinction I hold onto is between speed and haste. Speed is a short interval between adequate evidence and action. Haste is action before adequate evidence, usually driven by a deadline that someone invented. Speed is built. Haste is a symptom.