Home · Solutions · Sales & marketing
Solution · Sales & marketingThe guest asks in the evening and has the offer before breakfast, not after the weekend
Availability enquiries answered the same day
A robot reads enquiries from email and the website form, checks the availability calendar and drafts a complete reply with an offer: dates, room, price, deposit. The owner approves with one tap, or lets simple cases go out by themselves. After two days of silence, a polite follow-up goes out.
Executive summary
Availability enquiries arrive in the evenings and at weekends, when the owner is cleaning, cooking and handing out keys; the reply is written when there is a moment, which means many hours later, and meanwhile the guest books elsewhere.
The robot reads the enquiry, checks the calendar and drafts a reply with a concrete offer; the owner approves with a tap, and typical cases can go out by themselves; after two days of silence a follow-up goes out.
Response time falls from many hours to minutes by day and to morning after a night; more enquiries end in bookings, and the owner stops typing replies in an apron between duties.
the enquiry mailbox and form; the availability calendar; AI reply drafts; follow-up and a report in Microsoft Teams
Business problem
The guest does not wait for a reply; the guest books
An availability enquiry is the hottest moment in selling a stay: the guest sits with a phone, three tabs open, asking in several places at once. From that moment, time decides; accommodation market research says what common sense says: the first concrete answer wins.
Meanwhile, in a small guesthouse, replying competes with everything else about running the place: breakfasts, cleaning, key handovers, a broken boiler. Enquiries arrive in the evening and at weekends, exactly when guest work peaks. The reply appears many hours later and tends to be terse, written on the run: “we have rooms, welcome”, no price, no specifics, no deposit terms.
Every such delay is not just one lost booking; it is also the better class of guests, the ones who plan ahead and choose efficiently. What remains are last-minute enquiries and guests nobody else answered. The property slowly slides into dependence on commission portals, because there at least the booking happens by itself.
And yet the reply to an availability enquiry is, nine times out of ten, mechanical: check the calendar, compute the price, paste the deposit terms. That is work for a system, not for the owner at 10:40 p.m.
How it works today
Below is what the work looks like before anything is automated.
- PersonEnquiries arrive in the evening and at weekends, when guest work peaks
- WaitingThe reply waits many hours for the owner’s free moment
- Risk of errorThe guest with three open tabs books where the first answer came from
- PersonA reply written on the run lacks price and specifics; another email exchange follows
- Risk of errorA guest’s silence gets no follow-up; the thread dies
- Risk of errorBookings drift to commission portals, because there they happen by themselves
Why the current process costs more than it appears
The bill that never shows up in a budget.
- Every hour of delay lowers the chance of a booking; after a day the enquiry is usually dead.
- A guest lost to delay returns through a commission portal, if he returns at all.
- Replies without specifics breed email exchanges that cost further hours and further lost guests.
- The owner’s evenings spent at the mailbox are a personal cost no price list reflects.
Cost of inaction
The first row cautiously assumes every fifth enquiry dies of delay; in season, when guests ask in several places at once, it can be worse. You will know your own number after a month of measuring response times.
The second row is commission paid on bookings by guests who first asked directly and never got an answer; it is the bitterest line in this arithmetic.
A model organisation with realistic proportions – the numbers exist so you can run the same maths on your own data; they are not a client result.
A lakeside guesthouse: 11 rooms, a May-to-September season, the owner with one helper, enquiries by email and the website form, the calendar in a spreadsheet, Microsoft 365.
Around 170 enquiries a month in season; average response time 19 hours; no follow-up.
Evenings go on replying; some replies lack the price and deposit terms, breeding further exchanges.
A growing share of bookings from commission portals, because they answer for the property instantly.
The robot reads the enquiry, extracts dates and party size, checks the calendar and drafts the reply: rooms, the price for the stay, deposit terms, in the property’s tone; the owner approves with a tap from her phone, and simple cases governed by rules go out by themselves; after two days of silence a follow-up goes, and every conversation ends with a status.
In the modelled case response time falls to minutes during the day, and enquiry conversion rises by a quarter. Model numbers, not the guesthouse’s records.
Proposed solution
We start with your rules and tone: how you price (seasons, stay length, children, pets), your deposit and cancellation terms, how your hospitality sounds in an email. The AI receives those rules and examples of your best replies; from then on it drafts in that tone, not in a robot’s.
An enquiry from email or the form is read at once: the robot extracts dates, party size and side questions, checks the availability calendar and assembles a complete draft reply with a concrete offer. The owner gets it on her phone and approves with one tap, corrects with one sentence, or, for typical in-season cases, lets the rules send at once. The guest asking at 9:30 p.m. has the offer before closing his other tabs.
Silence after an offer does not stay silence: after two days a polite follow-up asks whether the dates are still wanted. Every conversation ends with a status: booked, declined, no reply, so for the first time the owner sees how many enquiries come, how many end in bookings and where the rest leaks away.
UiPath Orchestrator: the enquiry queue, follow-up schedules, retries and an audit trail; UiPath Integration Service connectors for Microsoft Teams and Outlook 365; AI models for reading enquiries and drafting replies
Price and deposit rules, the property’s tone learned from your replies, drafts approved from a phone, auto-send for typical cases, follow-up and a conversion report
The mailbox and form via Outlook 365; the availability calendar (a spreadsheet, a 365 calendar or a booking system via export or API)
How the automated process works
- AutomationThe enquiry is read at once: dates, party size, side questions
- AutomationThe robot checks the calendar and drafts the offer: room, price, deposit, in your tone
- PersonThe owner approves from her phone with one tap, or corrects with one sentence
- SystemTypical cases governed by rules go out by themselves; the guest has the offer in minutes
- AutomationTwo days of silence trigger a polite follow-up
- PersonEvery conversation ends with a status; the report shows conversion and leaks
Human-in-the-loop model
Automation handles
- Reading enquiries, checking availability and drafting offers in your tone
- Follow-ups after silence and statuses for every conversation
- The report: enquiries, response times, bookings, leaks
People decide
- Prices, terms and tone; the AI writes by them, not instead of them
- Approving offers, beyond the cases you yourselves mark as automatic
- Atypical conversations: groups, events, negotiations, special guests
Before and after
Systems and integrations
The stack is short on purpose: one engine, one execution layer, one place where a person decides.
Inputs
- enquiries from email and the form
- the availability calendar
- price, deposit and tone rules
- the owner’s decisions on drafts
Automation layer
- UiPath Orchestrator
- UiPath Robots
- UiPath Integration Service
- UiPath Action Center
Target systems
- sent offers with history
- follow-ups and conversation statuses
- the conversion report in Microsoft Teams
Human touchpoints: draft approvals from a phone; a weekly report; a price rules review before the season
Technologies used
the enquiry queue, follow-ups, retries, a record of every offer
Aextracting dates from emails written like humans write, and drafts in the property’s tone
Athe enquiry mailbox, approvals, reports
Aa spreadsheet, a 365 calendar or a booking system via export or API
Ba follow-up by text when the guest left only a phone number; on your gateway
BIllustrative economic model
Numbers you can check against your own data.
The model assumes a fast, concrete reply saves part of the enquiries, not all; some guests are lost for reasons nothing cures. Volumes and prices belong to the scenario; your own numbers go into the calculator alongside.
Run the maths on your data
An illustrative estimate based on your inputs. It models freed capacity, not promised savings.
Business benefits
- The guest gets the offer while still comparing, not after booking elsewhere
- Replies are complete: room, price, deposit; no email ping-pong
- The follow-up saves conversations that would have died in silence
- More bookings come direct, fewer through commission portals
- The owner gets her evenings back; what remains is a one-tap approval
The management view
- Selling stays stops depending on whether the owner happened to stand by the mailbox
- The property’s tone is written down and repeatable; the guest feels the hospitality in the first email
- Numbers on enquiries and conversion build the knowledge for price and season decisions
Board-level KPIs
time from enquiry to offer · enquiry conversion · conversations saved by follow-up · the share of direct bookings · evening hours at the mailbox
Security and governance
The automation has exactly the permissions it needs. Not one more.
- The robot works on enquiry data: dates, party size, contact; nothing beyond the offer’s needs
- Reply drafts are approved by a human, unless you yourselves mark a case as automatic
- Every offer and follow-up has a record: what, when, to whom
- The AI receives your rules and examples; it invents neither prices nor availability, and hands uncertainty to a human
- Data stays in your Microsoft 365 tenant; the robots run in the EU region of UiPath Automation Cloud
Why now
Guests compare several properties at once and decide within hours; response time has become the price of entry.
Portal commissions keep rising; every booking captured direct is pure profit on the same sale.
AI has learned to read emails written like humans write and to reply in a human tone; the automatic offer stopped sounding automatic.
Relevant executive roles
Sells even while handing out keys and sleeping; the evenings come back to her
Sees conversation statuses instead of verbal notes to “reply to the August lady”
Gets a concrete, warm offer in minutes and books without email ping-pong
Common questions and objections
The personal touch begins with somebody answering at all, while the guest is still waiting. The drafts are written in your tone, on your examples, and you approve each one until you decide the simple cases may go alone. The guest receives the same hospitality, only at 9:40 p.m. instead of the next day.
Seasons, stay lengths, extra beds, pets: those are all rules you apply in your head, and they can be written down. Where the rule ends, the draft arrives with an empty price and a question for you; the automat does not guess. Usually, once written down, 80% of enquiries turn out to be computable on the spot.
Which is why the source of availability is one calendar, the same one you synchronise with the portals. The robot reads from it, not beside it. If the calendars live separately today, putting them in order is the first implementation step, and it alone ends the double bookings.
When this is not the right solution
- A property fully automated on portals, with no direct enquiries: there is nothing to speed up
- No availability calendar at all: first one calendar, then offers from it
- Expecting the AI to negotiate prices and receive guests: it sells faster, not instead of you
A question for the next management meeting
How many hours did the last guest who did not book wait for our reply, and where does he sleep tonight?
Implementation approach
Scope without ambiguity, before anything is signed.
We deliver
- Written price, deposit and tone rules
- Automatic enquiry reading and offer drafts approved from a phone
- Auto-send for the cases you yourselves mark
- Follow-ups after silence and conversation statuses
- A conversion report and two weeks of parallel running
We need from you
- Access to the enquiry mailbox and the availability calendar
- Your pricing and deposit rules, plus a few best replies as tone examples
- A decision which cases may go out without approval
Stages
Discovery
Where enquiries come from, how many there are, what the response time is
Rules
Prices, deposit, tone; reply examples; the automat’s boundaries
Build
Enquiry reading, drafts, approvals, follow-ups, the report
Parallel run
Two weeks: drafts for approval, times and accuracy measured
Go-live
Auto-send for marked cases; a review before the season
A quick win. Effort depends on the availability calendar: one proper calendar is the shortest path; the AI learns your tone from your replies within an hour.
Tuesday, 9:30 p.m.: an enquiry about two rooms for a long weekend. The reply went out on Thursday. The guest was kind, said thank you, and sleeps by another lake.
Send us your last ten enquiries with your reply times. We return the arithmetic: how many bookings a year sit in the response time alone, and what your offer sent at 9:35 p.m. would look like.
Measure your response timeThe neighbouring process usually has the same problem
He called three times during opening hours. He booked where someone picked up.
See the solution Customer serviceOne inbox instead of four channelsEmail, Messenger, phone and a form. Four places where the customer could ask the same question.
See the solution Customer serviceGoogle reviews collected after every visitThe happy leave in silence. The unhappy one writes on Sunday at 11:40 p.m.
See the solutionIndustries where we deploy this most oftenSmall business & services