“I know my business.” The phrase is both true and dangerous. True, because twenty years of experience build valuable intuition. Dangerous, because that same intuition was formed in a market that no longer exists, contaminated by the biases of every human mind, and—most importantly—cannot be audited, transferred, or scaled. When a manager decides “by instinct,” the entire organization depends on an instrument no one else can use.
Making decisions with data does not mean replacing management judgment with spreadsheets; it means structuring the process so judgment operates on evidence rather than selective memory. This is the practical architecture.
The three levels of decision-making maturity
- Level 1 — Pure intuition: decisions are made through experience and conviction. Fast, indefensible, and non-repeatable. Quality depends on the decision-maker’s state of mind.
- Level 2 — Decorative data: the decision is made first and data is sought afterward to confirm it. The most common level in mid-sized companies—and more dangerous than level 1 because it produces the same decision with a false sense of rigor.
- Level 3 — Structured evidence: the question is framed first, data is examined methodically, alternatives compete, and the decision is documented with its logic. Auditable, transferable, and improvable.
The five-step structured decision process
1. Frame the decision as a question with options
“What should we do about sales?” is not decidable. “Should we hire another salesperson or invest that budget in demand generation?” is. A good decision question has concrete alternatives, a timeframe, and an implicit success criterion.
2. Define what evidence would decide—before looking at it
The ultimate anti-bias step: what data would make each option win? If hiring a salesperson wins when the bottleneck is service capacity, and demand generation wins when the bottleneck is lead volume, then the decisive data is current team utilization and pipeline coverage. Define it first, look at it afterward—the reverse order produces level 2.
3. Use the data you have, honestly assessing its quality
The usual excuse—“we do not have good data”—hides the fact that your CRM, accounting, and operations already contain more evidence than you use. Practical rule: for most management decisions, directional data available today beats perfect data that is never available. Clean interpretation matters more than precision—the interpretation biases are covered in interpreting customer data correctly.
4. Decide with judgment—the data informs, it does not govern
Data describes the past; decisions place bets on the future. Management judgment enters exactly here: weighing what data does not capture (context, relationships, strategic timing) and taking the risk with eyes open. The difference from pure intuition: now judgment knows what it is betting on.
5. Document the logic and schedule a review
Three lines are enough: what was decided, what evidence supported it, what we expect to happen, and when we will review it. This record turns every decision into a learning experiment—and builds the company’s strategic memory one decision at a time. It is the same principle as evidence-based marketing applied to general management.
Four questions that elevate any decision meeting
- What data supports this claim?
- What evidence would change our mind?
- What alternative are we rejecting, and why?
- When and with what number will we review whether it worked?
Once established as a management habit, these questions do more for decision quality than any software.
The minimum infrastructure
None of this requires a data department. It requires the business’s critical numbers to flow without heroics: an executive dashboard with essential KPIs—the design is covered in the KPIs every general manager should review—automatically fed from operating systems, supported by well-built commercial analytics.
Want to structure decision-making for your management team? Book a free diagnosis and let us assess your current decision-making maturity.