An advisor can lose an afternoon improving the wrong version of a trip. The proposal is coherent and the route works, but the feedback is a fog: less rushed, more special, not quite us. None of those phrases says which part should move.
The tempting response is another complete itinerary. That may produce more polish without producing more certainty. Now the client has two hotels, twelve reshuffled mornings and dozens of small differences to compare when the real uncertainty may concern one thing: pace, privacy, variety, cost or who gets the protected day.
Before moving a single pin, find the decision hidden inside the hesitation.
AI can help because it sorts and contrasts quickly. It should not decide what the client meant. The advisor chooses the question, removes false options, checks what is feasible and makes the recommendation.
The short answer
Do not ask AI for a replacement itinerary. Ask it to identify one unresolved choice that would materially change the proposal, then draft two genuinely different options with the cost of each made plain. Send that choice against the version the client has reviewed. Apply the answer only after the advisor has checked it.
Find the hinge, not the symptom
Feedback such as “too busy” is a symptom. The hinge is the choice that changes the structure of the trip.
Suppose the brief protects slow mornings but also contains several fixed-time highlights. Removing a restaurant will not resolve that tension. The useful question is whether the client wants to preserve the slow starts or keep more timed visits. Once that choice is clear, the smaller edits become easier.
A decision is worth taking back to the client when both answers would create a credible trip and the difference matters to how the journey will feel. If one option is clearly unsuitable, it is not a choice. It is a recommendation the advisor should explain.
- Set aside what is fixed. Dates, confirmed services and non-negotiable needs are not part of this choice.
- Name the protected outcome. Rest, depth, variety, privacy or price matters more than the individual item currently attracting attention.
- Write one unresolved question. It should change pace, route, scope or spend—not merely swap one similar attraction for another.
- State the consequence of each answer. The client needs to see what becomes easier and what becomes harder.
Let AI argue with the brief before it writes anything
This is a good moment for a bounded AI task. Give it the brief and the latest feedback, but forbid it from rewriting the itinerary. Ask it to separate client facts from advisor assumptions and to find the smallest choice that explains the uncertainty.
The first answer may still be wrong. That is useful. It gives the advisor something specific to reject: perhaps the supposed conflict is already settled in the notes, perhaps two preferences can coexist, or perhaps the real question belongs with a supplier rather than the client.
A polished answer is not the goal. A sharper conversation is.
A bounded pressure-test prompt
Do not create or rewrite an itinerary. From the brief and feedback below, identify one unresolved choice that would materially change pace, route, scope or spend. Separate client facts from advisor assumptions. Draft two feasible option shapes, explain what each protects and costs, and say “not enough information” if the notes do not support a fair choice. Do not recommend an option.
A hypothetical family, two honest shapes
Consider a hypothetical five-day proposal for a multigenerational family in Paris. This is not a real client case. The planning facts are deliberately spare: one long museum visit matters to an adult, a reliable afternoon pause matters to the group and the younger travellers want time that is not organised around formal sights.
The unresolved choice is not which museum to add. It is how concentrated the cultural time should be.
One option protects a single long museum morning and leaves the other mornings loose. It offers depth and fewer fixed starts, but that one morning asks more of everybody at once. The other spreads shorter cultural visits across three days. It reduces the longest block, but creates more scheduled mornings and more repeated movement.
Neither is the clever answer. Both have a cost. That is precisely why the family can discuss them.
| Option | What it protects | What it asks the family to accept |
|---|---|---|
| One deep cultural morning | Depth and flexible mornings elsewhere | One longer, more demanding visit |
| Several shorter visits | Shorter individual visits and more variety | More fixed starts and repeated movement |
Write a choice the client can answer in one breath
A useful client question could be: “Would you rather protect one long museum morning and keep the other mornings flexible, or schedule shorter cultural visits across three days?”
It names the difference without hiding the trade-off. It does not ask the client to compare the entire itinerary, and it does not dress the advisor’s preferred answer as the obviously sensible one.
Keep the options parallel. Change one meaningful variable at a time. If one version also changes the hotel, budget and evening plans, the response will not reveal which difference mattered. Give each option a plain title, one consequence and only the detail needed to decide.
The fairness test
Would the advisor be prepared to deliver either option well? If not, make a recommendation instead of manufacturing a choice.
Keep the answer attached to the proposal the client saw
Rooutie’s current proposal workspace gives this decision a better home than a loose email thread. First publish the client-safe version of the proposal. Decisions are tied to that version, so the question and response remain connected to what the client actually reviewed.
An advisor can ask the client to approve an item or scope, confirm information, or choose one or several options. The advisor can set who needs to respond and add an expiry when it is useful. Responses remain visible with the decision.
A satisfied decision does not silently rewrite the trip. It still needs Apply. If travellers choose differently, the disagreement remains an advisor conversation; the software does not turn it into a false consensus. Resolve the difference, select the intended outcome and only then apply it.
Before Apply, check the trip—not the prose
A client choice settles a preference. It does not prove that the chosen version works.
Return to the map and the day. Check that the option does not create awkward backtracking, remove required recovery time or disturb something the client already protected. Verify anything live or consequential with the responsible source: availability and terms with the supplier, transport with the operator, entry or access information with the venue, and official requirements with the relevant authority.
Then read any comment attached to the response. A client may choose an option and qualify it in the same breath. The useful outcome is not a green status beside the decision. It is an itinerary that reflects what the person meant.
Before applying the answer, confirm that:
- the proposal version is still the one the client reviewed;
- every option offered was genuinely feasible;
- the response came from the people whose agreement matters;
- any comment changes neither the meaning nor the scope of the choice;
- live facts have been checked with the source responsible for them.
Rehearse the question before using it with a client
Use this five-day Paris family itinerary on Rooutie as practice material, not as a route to copy. Write a clearly labelled hypothetical brief, identify one decision that would change the trip and prepare two options. If the exercise keeps producing a new itinerary instead of a clear question, the scope is still too wide.
More collaborative does not mean more versions
Client uncertainty does not always call for more creativity. Sometimes it asks the advisor to stop composing and name the choice.
AI can shorten that diagnosis. Rooutie can keep the answer beside the proposal. Neither can decide which trade-off is fair, which facts deserve another check or which recommendation the advisor is prepared to stand behind.
A proposal becomes collaborative when the client can see the decision that is truly theirs—and trust the advisor with the rest.
