CEO Decisions And Leadership

I learned this loop through decisions that looked reasonable on paper but exposed weak assumptions once real customers, employees, and cash flow entered the picture. The useful lesson was not to wait for perfect confidence. It was to make smaller commitments, name the risk, and decide in advance what evidence would change the plan. That discipline turns a lonely decision into a managed process.

The CEO Decision Loop

The CEO Decision Loop has five parts:

1. Frame the decision. Write the exact choice in one sentence, including the deadline and the result you need. “Should we open a second gym?” is too broad. “Should we sign a $6,000 monthly lease for a second gym by June 30 to add 120 active members within six months?” gives you something to examine.

2. Name the assumptions. List what must be true for the decision to work. For the gym, those assumptions might include 40 qualified leads each month, enough coaches to cover the schedule, and at least $18,000 in monthly membership revenue by month six. Separate confirmed facts from estimates. A signed lease quote is a fact; expected member demand is an estimate.

3. Choose the smallest useful test. Test the riskiest assumption before making the full commitment. The gym owner could run a four-week presale from the current location, offer 30 founding memberships for the new area, and spend $1,200 on local advertising. The test does not prove the entire expansion will work. It shows whether nearby customers will pay enough attention to justify the next step.

4. Set decision rules and a review date. Decide before the test begins what result will trigger action, caution, or a stop. For example: move forward with 20 paid presale memberships, revise the plan with 10 to 19, and stop below 10. Set a review date, such as July 8, and protect it on the calendar. Without a rule and a date, hope will replace judgment.

5. Revise the commitment. At the review, compare the result with the rule, identify what you learned, and choose one of three actions: continue, change the plan, or stop. Revision might mean a smaller location, different pricing, a later launch, or no expansion. The loop protects the business from defending a decision simply because the owner announced it.

The loop works because it matches the way uncertainty actually behaves. You rarely discover the full answer before acting. You create useful information by taking a controlled step, measuring the response, and limiting the cost of being wrong. A test also gives your team a clear reason for the work. Employees can understand “we are testing whether 20 local customers will prepay” more easily than “we are exploring growth.”

Use a simple decision page for every major choice:

| Field | What to record | |---|---| | Decision | The exact choice and deadline | | Desired result | The business result that matters | | Main assumption | The belief most likely to be wrong | | Test | The smallest action that checks it | | Measure | The number or behavior you will track | | Rule | The result that means continue, revise, or stop | | Review date | When you will make the next decision |

Do not fill the page with every possible concern. Rank assumptions by danger. A new service may depend on price, demand, delivery time, and staffing. If demand remains unproven, testing staff uniforms or software adds little value. Test the assumption that could damage cash flow first.

A Four-Week Test Before a Costly Expansion

Consider a plumbing company deciding whether to add an emergency evening service. The owner believes customers will pay a $175 call-out fee after 6 p.m. The company already receives about 22 evening inquiries each month, but only eight become jobs because the team cannot respond quickly. The proposed plan requires one technician on call, a scheduling service, and roughly $4,800 in monthly added cost.

The owner applies the CEO Decision Loop instead of immediately hiring two technicians.

1. Frame the decision on Monday. The question becomes: “Should we launch a paid evening service on July 1 if a four-week test produces at least $7,000 in collected revenue and keeps response time below 30 minutes?” The target leaves room for the $4,800 cost and a small contribution toward overhead.

2. Name the assumptions on Tuesday. The owner writes down three estimates: at least 40 customers will accept the call-out fee, the technician can complete 32 jobs in four weeks, and the scheduling service can answer every call within five minutes. The owner marks the first assumption as the biggest risk because free inquiries do not prove willingness to pay.

3. Run the smallest test for four weeks. The company adds a clear evening price to its website and voicemail, contacts the 22 recent evening inquiries, and offers service from 6 p.m. to 9 p.m. on Tuesdays and Thursdays. One existing technician volunteers for the limited test and receives a fixed evening premium. The company records inquiries, accepted fees, completed jobs, collected cash, cancellations, and response time.

4. Review the evidence on August 1. The test produces 31 paid jobs at an average collected invoice of $228, or $7,068 in revenue. The average response time reaches 24 minutes, but two jobs require refunds because the technician could not reach them in time. The result clears the revenue rule but exposes a staffing problem.

5. Revise the plan instead of declaring victory. The owner continues the service for another month, adds a second technician on Thursdays, and raises the response-time target to 20 minutes. The owner does not sign a long-term scheduling contract until the second test confirms demand and service quality. The expected outcome is better coverage without committing to the full $4,800 monthly cost before the process works.

This example shows why the loop needs both financial and operating measures. Revenue alone could hide late arrivals, refunds, exhausted staff, or unhappy customers. Choose measures that show whether the decision creates the result and whether the business can deliver it repeatedly.

Quick checklist

• Write the decision, deadline, and desired result in one sentence. - Separate facts from estimates. - Identify the assumption that could cost the most if wrong. - Run the smallest test that produces useful evidence. - Track cash, customer response, delivery quality, and team capacity. - Set continue, revise, and stop rules before you see the result. - Put the review date on the calendar. - Record what changed and what you will do next.

Mistakes That Turn Tests Into Guesswork

Mistaking activity for evidence

A busy test can still fail. Website visits, phone calls, meetings, and social media comments show attention, not demand. In the plumbing example, 60 people might ask about evening service while only 31 pay the fee.

Do this: Track the behavior that supports the business case: payment, signed agreement, repeat purchase, completed delivery, or retained customer.

Not this: Treat interest, compliments, or an employee’s confidence as proof that the decision works.

Changing the rules after seeing the result

Owners often set a target of 20 sales, reach 12, then explain why 12 should count. That move protects pride but weakens the test. If the original target no longer fits, record why and create a new test with a new rule.

Do this: Keep the original rule visible. If conditions changed, state the change, reset the test, and choose a fresh review date.

Testing too many variables at once

Changing price, service area, advertising, staffing, and delivery method together makes the result hard to read. A failed test may reflect poor pricing rather than weak demand.

Do this: Test the most important assumption first and hold the other major conditions steady when possible.

Waiting for certainty

A test cannot remove every risk. It should reduce the risk enough to support the next commitment. If the cost of a small test is $1,000 and the full decision costs $40,000, the test may offer valuable protection even when it produces an imperfect answer.

Do this: Decide what you need to learn, set the maximum test cost, and act when the evidence reaches that limit.

Your action now is simple: choose one pending decision that could affect cash, customers, or team capacity. Complete a one-page CEO Decision Loop before you discuss it with anyone. Write the main assumption, the smallest test, the success rule, and the review date. Then make the next commitment only as large as the evidence supports.

Strong organizations do not avoid uncertainty. They build a way to face it repeatedly, learn quickly, and lead with enough discipline to change course when the facts demand it.

End of chapter one. 33 more chapters in the full book.