GuardTrails
A feasibility check is a confidential report on a tour you have already designed or that someone has handed you. It is not there to decide in your place. It is there to complete your review: to show what you cannot see by eye on a programme of ten or twenty days, and to say where each signal comes from. The decision stays with you.
This page explains how to read that report. The eight checks it overlaps are set out in How to check a tour itinerary. For where the engine stands next to an operator or an AI assistant, see Operator, ChatGPT or engine.
What the report contains
The report has five parts, always in the same order.
1. The summary. A letter from A to D, a sentence that translates it, the number of blocking points and of points to check, the list of those points and the “Questions to ask” section. That section is always shown: “No question outstanding.” when there are none.
2. The corrected variant. What the engine would do with the same tour, on the same dates and over the same duration: only the differences are listed, each with its values before and after. This part has a study of its own: Overloaded tour: the corrected variant.
3. The findings by dimension. Six dimensions, each with what it found or “Nothing to report”:
- Route and driving time;
- Rhythm and nights per base;
- Season, climate, closures;
- Access and gateways (arrival, departure, location of each stage);
- Announced and estimated budget;
- Gaps and redundancies.
Each dimension also says whether it was measured in full, in part, or not at all, and why. That matters: “Nothing to report” on an unmeasured dimension does not mean “all is well”.
4. The gap to the market. The tour is set against comparable tours of the same zone, when the zone can be deduced: number of night bases, duration, nights per base, road hours per day, worst road day. For each indicator, the report says whether the tour is within the norm or above the market, or whether the indicator is not provided. When the zone cannot be compared, the report says so and calculates nothing.
5. The assumptions. Everything the tour did not say and the check had to assume, with the value kept and, where there are any, the other possibilities.
The grade: a rule, not an average
The letter follows a rule written in four lines:
- D: at least one serious finding;
- C: four findings to read or more;
- B: one finding to read, or a dimension that is unmeasured or partly measured;
- A: no finding, and every dimension measured.
“Accepted” findings and “over-estimate” findings do not count. There is no weighting and no average: a grade can be found again by hand by counting the points. It says how many things need a human eye; it does not say the tour is good or bad, and it does not replace your judgement on what you know of the ground and the client.
An example: the south-west coast of Portugal
The example comes from a tour in our corpus, run through the engine as part of our trials: fifteen days, four travellers, seven stages on the south-west coast of Portugal, from Porto Covo to Pontal da Carrapateira, with stays of one to three nights. It is not an operator's tour; it is a tour for which we can show the report.
The grade: B. “Feasible, a few points to check.” No blocking point, three points to check. The “Access and gateways” dimension is also unmeasured, for lack of stated arrival and departure: that too is what puts the tour in B rather than A.
The rhythm. Seven night bases for fifteen days. This is the point the report places in the gap to the market: the number of night bases (7) and the duration (15 days) are above the market; nights per base, road hours per day and the worst road day are within the norm. Read that way, the finding is precise: it is not the road that weighs, it is the number of bed changes. It is for you to say whether that is intended (a traveller who likes to move) or not (a group that would prefer two bases).
The access. One stage, Arrifana, could not be placed reliably: the report writes “not reliably located (a place of the same name was found, far from the other stages): legs not measured”. The engine would not measure journeys on a guess. The question to ask is stated in the finding itself: which town or address exactly?
The redundancies. The last stage (Pontal da Carrapateira) goes back to a place already left. The engine flags it; it does not say it is a mistake, since a return can be intended.
The assumptions. Four, written as they are in the report: “end date deduced”; “neither arrival nor departure stated: access journeys not counted”; “place not reliably located: legs not measured”; “comparison zone deduced from the stated region”. Each is written with the value kept.
What you take from it: three things to look at, one question to ask, four assumptions to confirm or correct. Nothing is decided in your place.
How to use it
Start with the assumptions. They are the most useful part and the quickest to read. If an assumption is wrong, the finding that depends on it changes: correct it before discussing the grade.
Read the “partly measured” dimensions as questions. When the report says a dimension could only be measured in part, or not at all, it says what was missing (activities not stated, closures not queried, neither arrival nor departure, no budget stated). That is the list of what the tour should have specified.
Read the gap to the market as an order of magnitude, never as a norm. “Above the market” means “different from what comparable tours do”, not “wrong”. A tour can be different on purpose.
Ask the report's questions as they stand. When data is missing, the report words the question rather than inventing an answer. It is made to be asked of the ground handler or the client.
Keep the decision. The engine flags and explains. You accept, refuse or adjust.
What the report does not say
- Facts. A stage can exist on the map and be closed or inaccessible: the engine does not know.
- Visas and formalities. Out of scope.
- Ferry, opening and crossing timetables. The engine detects that a constraint exists (a ferry, a mountain pass road, a permit), never whether it is compatible with a precise date.
- Bookings and availability. The report checks neither a lodging nor an activity.
- Price. The budget is an estimate on public scales, compared with the announced price when one is given; it is not a quote.
- Your knowledge of the ground. A point “to check” can be cleared in one sentence by someone who knows the place.
What we do not know
The grade is a rule written by us: another rule would give other letters. The thresholds are parameters of the engine, set for each case, not rules. The gap to the market depends on the comparable tours available for the zone and is not calculated when there are none. And a report on a single tour generalises nothing.
To have one of your tours checked
Fill in the beta form on /pro and tick “Check an existing circuit”. We are the ones who run the tour through the engine; the report is confidential and the tour is neither kept in a catalogue nor published. To place this check in your own profession, see For your profession and the page for ground handlers and operators.
Content of the report: description of the service, in beta. Example: tour from the GuardTrails corpus, engine trial.
A check looks wrong or incomplete to you? Write to us: corrections are published.