The inbox is where most travel businesses actually run their sales. A shared info@, a couple of forwarding rules, someone who keeps an eye on it, and a spreadsheet for the enquiries that get serious. It holds up for as long as every live enquiry still fits inside one person's head. Past that point it keeps working in the sense that mail still arrives, and stops working in the sense that anyone can tell you what is outstanding.
The leak is never dramatic. Nobody deletes an enquiry. A quote goes out on the Thursday before a long weekend and nobody picks it up again. A traveller replies from a different address and the thread splits in two. An enquiry for next April sits unread because it is not urgent, and by the time it is urgent, someone else has already sent the itinerary.
This is not a software shortlist. It is about what a travel pipeline has to hold from enquiry to repeat booking, why response time is your cheapest lever, how to follow up a quote across a decision that takes weeks, and where generic B2B CRMs fall apart.
Why shared inboxes break at travel scale
A single tailor-made booking runs several threads in parallel over several weeks: qualification, the first itinerary, a revision, a lodge availability check with a DMC, the deposit invoice, insurance, final documents. In a shared inbox the traveller thread and the supplier threads sit side by side with near-identical subject lines, and nothing in the system knows which of them is waiting on you.
The consideration window is what really breaks the model. Expedia Group's Path to Purchase research, published on its partner blog, found that travellers view 141 pages of travel content in the 45 days before booking, spending just over five hours with it. Your quote has to survive that entire window, and an inbox has no memory of week two.
Enquiries also arrive on channels an inbox cannot hold: web form, WhatsApp, a call taken in the car, an Instagram DM. And an inbox has two states, read and unread. Unread is not a stage. It cannot tell you that a quote went out eleven days ago and nobody has spoken to that traveller since. The spreadsheet is meant to bridge that gap, and it does, right up until the week you are busiest, which is the same week the enquiries peak.
Travel runs on four clocks, not one
Here is the structural point most CRM decisions get wrong. A generic sales tool has one date that matters: the close date. Travel has at least four.
The enquiry date is what you measure marketing against. The quote date is what your follow-up runs off. The departure date drives operations, staffing and supplier commitments. The return date starts the clock on repeat marketing. A March booking for a November departure is a March cash event, a November delivery event and a January-next-year marketing event. Store one of those and you will end up marketing off the wrong one.
With that in mind, write down the stages a booking genuinely moves through in your business:
- New enquiry, with source, market and campaign captured at submission
- Qualified: dates or flexibility, party composition, budget band, trip type, decision timeline
- Quote sent, with version number, date, price, margin and a validity date
- In revision, because version three is normal in tailor-made, not a warning sign
- Deposit paid, your only honest definition of won, then balance due
- Pre-departure, then travelling, where a change request is an operations job, not a sales job
- Returned, which triggers the review and referral ask, then nurture for a trip two years out
Response time is the cheapest lever you own
The best evidence on response speed is not from travel and it is old, so treat it as directional. In The Short Life of Online Sales Leads, published in Harvard Business Review in March 2011, James Oldroyd, Kristina McElheran and David Elkington audited 2,241 US companies by sending each a web-generated test enquiry. 37% responded within an hour, 24% took more than 24 hours, and 23% never responded at all. Among those that did reply within thirty days, the average wait was 42 hours.
In a separate study of 1.25 million leads across 42 US companies, the same authors found that firms attempting contact within an hour were nearly seven times as likely to qualify the lead, which they defined as having a meaningful conversation with a key decision maker, as those trying an hour later, and more than 60 times as likely as those waiting 24 hours or longer. One of the causes they name is pulling leads out of CRM databases daily rather than continuously, which is precisely what a web form emailing an inbox somebody checks twice a day does.
That research is not about travel, so take the mechanism rather than the multiple. Someone planning Kenya in August is not enquiring with you alone. Comparison is the entire point of an enquiry form, and the shortlist gets built in one sitting. The first operator back with something substantive sets the frame: the route, the price anchor, which properties are worth paying up for. Everyone who follows is answering your itinerary rather than proposing their own, and that advantage costs you nothing but process.
So: acknowledge automatically within minutes, get a substantive human reply out inside four working hours, and set an out-of-hours rule for your biggest source markets. If a third of your enquiries come from the US, your 09:00 CET is their 03:00, and an enquiry sent at their bedtime is competing against operators who answered it before you woke up.
Quote follow-up is a sequence, not a nudge
The usual pattern is a quote, one just checking in five days later, then silence. Nothing about a several-thousand-euro decision resolves in five days. The traveller is coordinating with a partner, checking annual leave against school holidays, and waiting on two other operators to come back to them.
An illustration with invented numbers, to show the shape rather than to promise a result. Say you send 40 quotes a month at an average trip value of EUR 4,200 and convert 22 per cent: 8.8 bookings, EUR 36,960 of trip value, about EUR 5,544 of margin at 15 per cent. Move conversion to 26 per cent, which is between one and two extra bookings a month that would otherwise have gone quiet, and you get 10.4 bookings, EUR 43,680 and EUR 6,552 of margin. That is a little over EUR 12,000 of additional margin across a year, from emails you should have been sending anyway. Run the same arithmetic on your own average trip value and margin. The higher your ticket, the less conversion has to move before the sequence pays for itself.
Note what is absent from the sequence below: invented urgency. In travel the honest constraints are real ones, such as camp space you are genuinely holding, a seasonal rate change or a permit deadline, and travellers spending this much can smell a fake one from a long way off.
- Day 0, within minutes: an acknowledgement naming who is handling it and when
- Day 0 to 1: the human reply, either the quote or two or three qualifying questions
- Quote plus 2 days: anything unclear, offering a call and naming a trade-off
- Quote plus 5 days: value not pressure, such as a lodge alternative at a lower price
- Quote plus 12 days: the honest constraint, such as camp space you hold or the price expiry
- Quote plus 25 days: the graceful close, asking whether to park it, which gets replies
- Then nurture, re-approached at the right seasonal moment for their trip
Tag for the questions you will actually ask
Tag backwards from the report you want. If you cannot name the decision a field informs, do not create it. The pull in travel is to record everything a traveller tells you, because on a tailor-made trip it genuinely is all relevant, but trip notes and reporting fields are two different jobs and only one of them belongs in a dropdown. Six fields your consultants complete reliably beat forty they complete sometimes: a field that gets filled in only when someone remembers does not give you a partial answer, it gives you a report nobody trusts.
Two earn their place above the rest. Lost reason is the field nobody completes and the one most likely to change your pricing. And channel attached to the booking rather than to the enquiry is what shows you whether paid search is bringing volume while a destination guide brings the fourteen-day trips, a distinction that stays invisible for as long as you only count enquiries.
- Trip type: safari, self-drive, trekking, honeymoon, family, small group departure
- Source market and language, kept separate
- Party composition, not headcount: two adults, or two adults and two children under 12
- Travel month, and whether dates are fixed or flexible
- Value band, with per-pax value stored separately from total
- Acquisition channel down to campaign, attached to the booking
- Lost reason from a fixed list: price, dates unavailable, chose another operator, postponed, went quiet
Forecast booked revenue and pipeline as two numbers
Booked revenue means the deposit is paid. Pipeline means quotes are out, weighted by stage. Report them separately and never add them together, because the moment they become one number nobody can tell you what is real, and in travel the distance between the two is measured in months rather than days.
Report both by booking month and by departure month. Booking month is cash and marketing efficiency. Departure month is capacity, staffing, guides and supplier exposure. An operator with a wonderful March of sales, all of it departing in one week of October, has a problem the P&L will not reveal for seven months.
Forecast on margin, not trip value: a twelve thousand euro tailor-made with flights routed through you can carry less margin than a four thousand euro land-only. And compare like months year on year, because month-on-month movement is close to meaningless in a business where enquiries pile up in January and the shoulder season looks like a collapse when it is simply April.
Where generic B2B CRMs break in travel
Plenty of travel businesses run perfectly well on a general-purpose CRM. But there are places where the data model underneath does not match how travel actually works, and it is worth knowing which ones you are choosing to live with.
- One deal, one contact. Travel has a booker, a payer and travellers who are neither; passports and dietary needs belong to people, not to a deal.
- No departure date. Close date is the only date the object understands, so operations ends up living somewhere else entirely.
- No per-pax anything. Occupancy, single supplements, child rates and group tiers cannot live in a field called amount.
- Revisions look like regression. Version three is healthy in tailor-made, but a CRM reading a reopened deal as a step backwards produces reports your team quietly ignores.
- No supplier side. Half the correspondence on a booking is with a DMC, a lodge or a guide, and if there is nowhere to put it, it stays in the inbox.
- Sales-cycle assumptions borrowed from software. A tool that flags anything stale after 30 days will bury exactly the enquiries you most want, because a trip booked eight months out can sit quiet for weeks and still be entirely live.
- Compliance blindness. Under the EU Package Travel Directive 2015/2302, organisers must give specified pre-contractual information on the standard form in Annex I, and Article 8 places the burden of proof for having done so on the trader. If your system cannot show what was sent and when, that is your evidence gone.
The mistakes I see most often
Migrating everything. Five years of dead enquiries loaded in on day one guarantees the team stops trusting the data by week two. Move live records only.
Buying a tour operator platform and calling it a CRM. Itinerary builders produce beautiful documents and are poor at telling you which five quotes need chasing on Thursday. The gap between the two is where a travel-native CRM sits, including our own, OpenVoy.
Building a reporting tool the consultants will not use. If updating a stage takes more clicks than replying to the email, they will reply to the email.
Leaving the shared inbox running alongside. Two systems means neither is true. Point the form at the CRM and forward the inbox in.
Treating a verbal yes as won. In travel the deposit is the event, and a pipeline of verbal yeses forecasts a quarter that does not arrive.
Skipping lost reasons. Without them you guess, and the guess is almost always price, which is convenient because price is the one answer that asks nothing of you. Speed, itinerary quality and follow-up never get their turn.
Never marketing to past travellers. Someone who did your Vietnam trip two years ago costs you nothing to acquire and already knows how you work. Check what consent basis you actually hold for past travellers before you send anything.
What to migrate first, and in what order
Do this in stages small enough that each one is useful on its own, and run it outside your peak enquiry weeks.
Week one is capture. Point every enquiry form, your WhatsApp number and phone enquiries at one place, and capture source, campaign and source market at submission. You will learn how many enquiries you actually receive, which is rarely the number you thought, and how many of them land outside your office hours.
Week two is stages. Six, no more: new, qualified, quote sent, deposit paid, travelling, returned. Six that everyone can recite beat twelve that nobody updates.
Week three is response time. Get the acknowledgement live, put an owner on every enquiry within the hour, and measure time to first reply split by source market, so you can see whether your US and Australian enquiries are quietly waiting overnight. Measuring it moves it.
Week four is the sequence. Build the touches, automate the reminders, and keep the emails human, because a tailor-made quote followed up by something that reads like a template undoes the itinerary it is chasing.
Month two is fields, on live enquiries only, and do not backfill. Month three is reporting: booked versus pipeline, by booking month and departure month, in margin, with cohort conversion and lost reasons.
The goal is not a complete database. The goal is that on Monday morning one person opens one screen and sees every quote that needs a reply this week, and no traveller waits until Tuesday because it was somebody else's turn.