What Happens When It Goes Wrong

Every operator's website has a paragraph about being there when you need us, and none of them say what they mean by it. We would rather set out the actual failure modes, what we actually do about each, and — this is the part that makes the piece worth reading — what we cannot do.

The reason for writing it is simple. The decision a guest is really making when they choose an operator is not about itineraries, because everybody's itinerary looks similar on paper. It is about who they want on the end of the phone at two in the morning when something has gone wrong in a country where they do not speak the language. That is an unglamorous thing to sell and it is the whole proposition.

So here is our list. It is not short, and some of the entries are not flattering to us.

Somebody gets ill

This is the most common thing by a distance, and most of it is minor. A stomach upset on day three is so routine that we plan around it rather than react to it: the second and third days of a trip are deliberately the least demanding, because that is when it tends to happen.

What we do is send a doctor to the hotel rather than send the guest to a hospital. A doctor's visit to a good hotel in any major Indian city is a straightforward thing to arrange, usually within an hour or two, and it is both faster and less frightening than a hospital waiting room. We pay for it and settle it later rather than having a guest stand at a counter feeling awful.

Where it becomes serious — a cardiac event, a fall, a reaction to something — the honest position is that India's private hospitals in the major cities are genuinely excellent and the smaller towns are not. A guest taken ill in Udaipur or Jaipur is in a good place. A guest taken ill in a tiger reserve at six in the morning is three hours from anywhere, and that is the real reason we are careful about who we send on remote itineraries and when.

What we cannot do is make your insurance pay. We will get you treated and we will deal with the hospital, but if you have travelled on a policy that excludes a pre-existing condition you did not declare, or that caps medical evacuation below what an air ambulance costs, that is a conversation that happens at the worst possible moment. We ask every guest to check two things before they travel: the medical evacuation limit, and whether their existing conditions are declared. Almost nobody does. It is the single most consequential thing a guest controls.

And the thing we insist on, which some guests find excessive: we want to know about conditions, medications and allergies before the trip, not during it. Not to be intrusive, but because a guide who knows that a guest is diabetic plans the day differently, and a hotel that knows about a severe nut allergy handles the kitchen differently.

A flight is cancelled

Domestic Indian aviation is better than its reputation and still cancels flights, usually for fog in the north between late December and January, occasionally for weather in the hills, and sometimes for reasons nobody explains.

What we do is rebook before the airline's queue forms, which is a matter of being awake and watching rather than being clever. The practical difference between a good operator and an indifferent one on a cancellation day is about forty minutes, and forty minutes is the difference between the last two seats on the next flight and a night in an airport hotel.

Where there is no flight, we drive. North Indian distances are long but not impossible, and a car leaving at four in the morning beats a flight at noon more often than people expect. We would rather put a guest in a good car with a driver we know for seven hours than leave them in a terminal.

What we cannot do is conjure capacity in peak season. Between Christmas and the first week of January, and around the big festivals, flights are genuinely full, and a cancellation on 28 December is a harder problem than the same cancellation in August. That is one of the reasons we are blunt with guests about what peak-season travel actually costs in flexibility as well as money.

The structural answer is the one nobody wants to hear: we build the itinerary so that a lost half-day does not destroy anything. If a trip only works when every connection works, it is a badly designed trip. That means fewer places and longer in each, which is an argument we make for other reasons too.

A road closes

Strikes, protests, political rallies, processions, landslides, floods, and the occasional bandh that closes a whole state's roads for a day. These happen and they are rarely announced far enough ahead to plan around.

What we do is know about them first, which is the only real advantage available. Our people are local, they hear about a bandh the evening before rather than on the morning, and the practical response is to move the day's schedule rather than cancel it — see the thing at the other end of town, or do tomorrow's plan today.

What we cannot do is move a guest through a closed road. Nobody can, and an operator who tells you otherwise is going to put you in a car with a driver who takes a risk on your behalf. We do not do that, and we have had the conversation where a guest very much wanted us to.

The honest admission here is that the Himalayan itineraries carry weather risk that cannot be designed away. A road to Ladakh or Spiti closes when it closes. We tell guests this before they book rather than after, and we build those trips with a contingency day in them. If the contingency day is not used, it becomes a good day somewhere.

The thing you came to see is shut

Monuments close for restoration, for VIP visits, for religious occasions and sometimes for no stated reason. The Taj Mahal closes on Fridays, which catches more itineraries than anything else on this list. Temples close for hours in the middle of the day. A fort can be shut because a film is being made in it.

What we do is check the week before and again the day before, and we have a second plan for every major site. That sounds obvious and it is not universal.

What we cannot do is reopen it. When a guest has come to India principally for one building and that building is closed on the only day they are in the city, that is a failure, and the question is whether it was foreseeable. A Friday at the Taj is foreseeable and if we get that wrong it is our fault. An unannounced closure for a state visit is not, and in that case what we owe the guest is a real alternative and an honest explanation, not a cheerful pivot.

We will also say what we do not do, because it goes on elsewhere. We do not pay anyone to open a closed monument, and we do not arrange access that depends on somebody breaking a rule. It is not a moral posture so much as a practical one: those arrangements fail at the worst moment and they fail with the guest standing there.

The hotel has overbooked, or is not what was promised

It happens, including at properties whose names would surprise you, and it happens most in peak season and around weddings.

What we do is find out before the guest does. The check-in is confirmed the day before, not on the day, and if a room has been given away we would rather have that fight at four in the afternoon than at eleven at night with a tired guest standing in a lobby.

Where a property cannot honour a booking we move the guest up rather than sideways, and we absorb the difference. That is not generosity, it is the only defensible response to a failure that is ours in the guest's eyes regardless of whose fault it actually was.

The harder case is a hotel that is open, has the room, and is simply not as good as it was. Properties decline. A hotel we loved two years ago can change hands, lose a general manager, and become ordinary, and the guest finds out before we do unless we are paying attention. We have pulled properties from itineraries for exactly this reason, and we have written elsewhere about doing it mid-season when something we could no longer stand behind came to light.

What we cannot do is pretend a hotel is fine when it is not. If a guest tells us a property has disappointed them, the useful response is to move them, not to explain why they should be happy.

The guide is wrong for the guest

This is the failure we take most seriously and the one least discussed, because it does not look like a failure. Nothing breaks. The guest simply has a flat day with somebody who is knowledgeable and not right for them.

Guides are not interchangeable. A guide who is superb with a family of four is not the right person for a professor of Mughal architecture, and the reverse is more often true. The person who can hold a nine-year-old's attention for two hours in a fort is doing something genuinely difficult that has nothing to do with how much they know.

What we do is match deliberately, and change without argument. If the first morning does not work, we change the guide for the afternoon. There is no face to be saved here and no conversation to be had about whether the guest is being fair. We would rather lose the margin on a day than have somebody endure three more of them out of politeness.

What we ask in return is to be told. Guests are often reluctant, because the guide has been perfectly nice and nobody wants to get somebody into trouble. Nobody gets into trouble. Tell us on the first evening, not in the feedback form.

Something happens at home

A death, an illness, a crisis that means the trip ends now. It is the rarest of these and the one that matters most.

What we do is get the guest to an airport and onto a flight, and deal with everything else afterwards. Nobody settles a bill or signs anything in that situation. We cancel what can be cancelled, we argue with suppliers about what cannot, and we send the account later — and in practice we often do not send one at all.

What we cannot do is make a prepaid non-refundable rate refundable. We can and do ask, and more often than people expect a good hotel will waive it. But we will not promise a guest in that state that nothing will be lost, because that is a promise made by somebody else.

This is where travel insurance matters again, and for a reason guests rarely consider: cancellation and curtailment cover is usually the larger number on the policy, not the medical one.

The things that are our fault

It would be dishonest to write a piece about failure that is entirely about weather and airlines.

A car arrives late. A driver takes a wrong turn and a guest misses the light they got up at five for. A hotel is booked for the wrong night. A reconfirmation is not made. A guest's dietary requirement does not reach the kitchen. A special occasion is not marked. These are ours and they happen.

What we try to do about them is unglamorous: fewer moving parts, confirmed the day before, by people who work for us rather than for a platform. Most operational failure in this industry is a handover failure — the information existed and did not travel. The answer is not better intentions, it is fewer handovers.

When it does happen, we say so. We have found that guests forgive almost anything that is explained honestly and quickly, and forgive very little that they discover themselves. The instinct to manage the information is the instinct to avoid, and it costs far more than the original error.

And we do not haggle about remedy at the end of a trip. If something was ours, we take it off the bill without a conversation about proportionality. The alternative — a dispute by email three weeks later — is worse for everybody and cheaper for nobody.

What actually determines how bad it gets

Almost none of this is decided in the moment. By the time something has gone wrong, the outcome is mostly already fixed by three things settled long before.

The first is slack. An itinerary with no give in it converts a small problem into a ruined week. This is the real argument for fewer places and longer stays, and it is an argument we make constantly for aesthetic reasons when the practical reason is sturdier: a trip with room in it survives things.

The second is who answers the phone. Not whether somebody answers — everybody has a number — but whether the person answering can make a decision. A twenty-four-hour line that takes a message and promises a callback is a worse thing than no line at all, because it costs the guest the hour in which the problem was still small.

The third is whether the people on the ground work for the operator. A trip assembled from a marketplace of local suppliers looks identical on the itinerary and behaves completely differently when something breaks, because nobody in the chain has any reason to absorb a cost on behalf of a guest they will never see again. This is the single biggest difference between operators and it is invisible until the day it matters.

We would add a fourth, which is about the guest rather than us. Tell somebody early. The problems that become serious are almost always the ones a guest sat on for a day out of politeness — the stomach ache on Tuesday that becomes a hospital on Thursday, the guide who was not right on Monday and is still not right on Wednesday, the room that was wrong on arrival and was endured for three nights. There is nothing to be gained by being a good sport about it, and the cost of mentioning it is nothing.

Why we published this

Because the alternative is a sentence about dedicated support, and because we think a guest deciding where to spend a serious sum is entitled to know what the failure modes are before they are in one.

And because writing it down is a commitment. Everything above is what we do; having set it out in public, we can be held to it. That seems to us a reasonable price for asking somebody to trust us with three weeks of their life in a country they do not know.