Critical thinking makes a decision easier to explain: what do we know, what are we assuming, and what would change our minds? Creative problem solving expands the possible responses before the team commits to one. Use both when a recurring business problem has several plausible causes.
Write the problem without choosing its cause
Start with an observable gap, a time period, and the people affected. “Our team needs a new scheduling app” already chooses a solution. “Six of 20 customer installations missed the promised date last month” gives the team something to investigate.
Bring actual case records, the promise made to the customer, and the people who perform the work. Ask when the pattern began, which cases succeeded, and what differs between successful and delayed cases. Give one person responsibility for the decision and state the deadline. Scale the investigation to the consequences and reversibility of the choice.
Separate observations, explanations, and missing evidence
In this fictional installation team, four of the six late jobs lacked a confirmed part before scheduling; two had all parts but no available installer. That pattern suggests two issues worth checking. It does not establish that every delay has the same cause.
On a small screen, scroll the table sideways to read all columns.
| Statement | Classification | Next check |
|---|---|---|
| Six of 20 jobs were late | Observation from job records | Verify the promised and actual dates for all 20. |
| The warehouse is understaffed | Possible explanation | Compare part availability, stock errors, and order timing. |
| Confirming parts first will reduce delays | Testable assumption | Pilot the check and record both delays and added work. |
Ask someone to look for a case that contradicts the preferred explanation. Review successful jobs as well as failures. A compelling anecdote can reveal a mechanism; the case log shows how often that mechanism actually appears.
Generate alternatives, then compare them on stated criteria
Have each participant write options independently before discussing them. Consider changing the sequence, removing an unnecessary handoff, improving information, and changing capacity. Invite the person who will maintain the solution to identify operating costs and failure cases.
NASA's decision-analysis guidance separates the decision, criteria, alternatives, evaluation, and recorded recommendation. It also calls for documenting uncertainty. For a small business decision, borrow that discipline and keep the comparison proportionate.
The team sets two mandatory conditions: keep customer data in approved systems and avoid promising an installation before readiness is confirmed. It then compares the remaining options using the same practical questions.
On a small screen, scroll the table sideways to read all columns.
| Option | What it may improve | Cost or unresolved assumption |
|---|---|---|
| Check parts before offering a date | Addresses the four jobs with missing parts | Adds checking time; stock records may be wrong. |
| Reserve an installer slot earlier | Addresses scheduling availability | May leave capacity unused while parts arrive. |
| Purchase new scheduling software | Could coordinate parts and people | Requires evidence that integration and adoption solve the observed problems. |
Keep uncertain entries visible. If you use weighted scores, define the scale and test whether modest changes in weights reverse the result. A precise-looking total should lead back to reasons and evidence.
Run a small test with a stopping rule
ASQ's Plan-Do-Check-Act description treats improvement as a cycle of planning, testing, checking, and acting on what was learned. Decide whether to continue, adjust, or stop based on the whole result, including added work and customer delay.
Leave a decision that someone else can inspect
Record the selected option, evidence, remaining assumptions, owner, review date, and conditions that would trigger a change. Keep rejected options and the reasons for rejecting them. This makes later revision easier when a supplier, workload, or constraint changes.
Use the decision worksheet to capture one decision. AI can suggest alternative explanations or summarize a de-identified case log; a person should verify every asserted fact and arithmetic result before it influences the decision.
For a problem involving queues and handoffs, continue with Business Process Management. For the people skills needed to run the discussion, see Leadership Development.