Good recommendations show options
When the decision matters, present a small number of viable options and explain the trade-offs. A recommendation should be connected to the client's priorities.
| Question | Practical answer | |
|---|---|---|
| Option A | Fastest short-term fix | Lower immediate cost; may preserve technical debt |
| Option B | Targeted improvement | Moderate effort; addresses root cause in a bounded way |
| Option C | Platform / architecture change | Higher initial effort; may improve scalability and reduce recurring constraints |
Use decision criteria
- Business value
- Time to value
- Total cost
- Risk
- Scalability
- Operational fit
- Reversibility
Not every criterion needs a numerical score. Often a clear qualitative comparison is enough to expose the decision.
Recommendations become stronger when the trade-off is explicit
Clients do not always need four options. They need enough contrast to understand the decision. One option may be fastest, another more scalable, another cheaper but dependent on manual work. Put those differences on the table.
Use criteria before preference
Criteria might include business fit, delivery time, cost, maintenance effort, scalability, reversibility and integration risk. A simple decision matrix forces the team to explain why an option scores well instead of relying on “this feels better.”
The recommendation should still contain judgment. A matrix is a way to expose the reasoning, not outsource the decision to a spreadsheet.
Compare three solutions for a client who needs subscription capability. Choose five criteria, define their relative importance, and write a short recommendation that acknowledges the biggest trade-off.