It appears in no cost centre, has no owner and generates no data. Which is precisely why it never gets funded - and how to make it visible enough to defend.
Every finance function in travel can tell you the cost of its contact operation to two decimal places. Almost none can tell you the cost of the contacts it did not handle. The first is a cost centre with a budget line and an owner. The second is a revenue event that did not happen, recorded nowhere, attributable to no one.
This asymmetry has a predictable consequence. Proposals to reduce the cost of handling contacts get funded, because the saving lands in a line somebody owns. Proposals to reduce the number of contacts you fail to handle do not, because the benefit lands in a line that does not exist.
It applies wherever enquiries arrive and people answer them: a hotel group, a tour operator taking departure questions, a travel agency, an attraction handling group bookings, a transport operator fielding disruption calls. The shape of the problem does not change with the segment. Only the value of the enquiry does.
Why finance cannot see it
Three structural reasons, and it is worth being precise about them because each has a different fix.
· There is no line item.Lost revenue from an unanswered enquiry is not a cost, so it cannot appear as one. It reduces revenue against a counterfactual, and management accounts do not carry counterfactuals.
· There is no owner.Reservations or sales owns conversion on the contacts it handles. Nobody owns the contacts that never reached a human. In most operators this sits in the gap between commercial and operations, which is the gap where things go unfunded.
· There is no data.Legacy telephony often cannot distinguish an abandoned booking enquiry from a supplier calling twice. Where the data exists it usually sits in a system finance has never seen a report from.
Building the number
A defensible model needs four inputs. The discipline is to source each from somewhere the business already trusts rather than from a vendor's assumption or an industry average. There is no published benchmark for this figure in European travel, in any of the major markets. We went looking. It does not exist, which is exactly why the model has to be built from your own numbers.
1. Unanswered and abandoned inbound volume, by hour and by channel. Source: your telephony or contact platform reporting. If it cannot produce this, that gap is itself a finding worth putting in the paper.
2. The proportion of inbound contact that is a booking or sales enquiry rather than service, supplier or internal. Source: a sample of handled contacts, coded manually over a representative fortnight. Two weeks of manual coding produces a far more defensible ratio than any benchmark.
3. Conversion rate on comparable handled enquiries. Source: your own reservations or sales reporting. Use handled-enquiry conversion as the proxy and state that assumption explicitly.
4. Average transaction value for the relevant segment. Source: your own revenue reporting, segmented - a blended average will overstate or understate depending on which enquiries you are missing, and a CFO will spot it.
Multiply through and you have a first-order number. Present it as a range rather than a point estimate, with the conversion assumption as the variable, because that is the input a finance reader will challenge first and offering the range pre-empts the challenge.
The second-order costs, which are usually larger
The first-order number understates the position, often substantially, because three further costs sit behind it.
· Wasted acquisition spend.You paid to generate the enquiry. Every unanswered booking call arriving from a paid channel is acquisition spend converted into nothing. This cost is real, already incurred, and sits in a budget line marketing does own - which makes it the easiest of the three to evidence.
· Commission on the recapture.Someone who cannot reach you directly frequently buys from you anyway, through a third-party platform. You do not lose the sale; you lose the margin, and you pay a commission on a customer you had already acquired. This is the cost that most changes the shape of the business case, because it converts a speculative lost-revenue argument into an actual, invoiced expense.
· Service recovery.Some proportion of unanswered enquiries return later as complaints, escalations or negative reviews, each of which has a handling cost and, in a review-driven industry, a downstream conversion cost.
Three objections a CFO will raise
Anticipate these in the paper rather than in the meeting.
· The counterfactual."You are assuming every unanswered call was a booking we lost." Correct, and you are not. State the enquiry-proportion ratio, state its source, and show the model at a conservative, central and optimistic conversion assumption. A model that survives its own pessimistic case is far more persuasive than one that only works at the optimistic one.
· They called back."They rang back." Some did. Your own repeat-caller data tells you what proportion - the same number attempting several times is a recovery, distinct numbers attempting once each are not. If your system cannot distinguish these, say so and treat it as a range rather than guessing.
· Where does it land."If we fix this, where does the money actually show up?" This is the fair question and the one that decides funding. The answer needs to be specific: direct booking share, commission expense as a percentage of revenue, and conversion on the assisted channel. Name the lines, name the owners, and agree the measurement before the investment rather than after.
What a defensible model looks like
It uses your data, not a benchmark. It states every assumption in a cell a reviewer can change. It presents a range. It separates the first-order revenue loss from the second-order commission and acquisition costs, because the second-order costs are more certain and land in lines that already exist. And it names, in advance, the three reported numbers that should move if the investment works.
Built that way, the proposal stops being a request to fund an unmeasured problem and becomes a request to stop paying a measurable one. Those are very different conversations, and only one of them gets funded.
See the worked example: what an unanswered enquiry costs, calculated step by step