A stakeholder asks "which option is better?" for a decision with genuine trade-offs. How should an architect answer?

The answer that fails

"It depends." True, unhelpful, and it transfers the analytical work back to the person who asked. It is also frequently a way of avoiding the responsibility of a recommendation.

The answer that works

"Better for what? Here is what each gives you and what it costs, here is my recommendation, and here is what would change my mind."

That structure does three things: it makes the criteria explicit, it gives the stakeholder the ability to disagree on grounds rather than on authority, and it commits the architect to a position without pretending there is no trade.

The components

1. Name the dimensions the options differ on. Not all of them — the two or three that actually decide it. Latency, cost, time to deliver, operational burden, reversibility.

2. State each option's position on each dimension, concretely. "Option A delivers in six weeks and costs 40% more to run; option B takes twelve weeks and is cheaper to operate and easier to change later."

3. Identify which dimension dominates, and say why. This is the architectural judgement, and it is where the value is. Usually one dimension matters far more than the others given the business context, and naming it collapses the decision.

4. Recommend, with the reasoning visible, so it can be challenged on the reasoning rather than on authority.

5. State what would change the recommendation. "If we expect to enter three more markets next year, B becomes clearly right." This is what lets a stakeholder who knows something you do not correct the analysis.

The trap to avoid

False balance. Presenting two options as equally weighted when one is clearly better in context is not neutrality; it is abdication. Stakeholders ask architects because they want judgement, and an architect who withholds it is not being rigorous.

The related trap

Hiding a recommendation inside apparent analysis, where the options are framed so only one is viable. It is detected, and once detected it costs the trust that makes future recommendations acceptable.

Present the options fairly and recommend openly. That combination is what makes an architect's judgement worth asking for.