Independent hotels, estates and wedding and events venues: the first place we’re building AI revenue systems. Years of enquiries, bookings and guest history, spread across systems that don’t talk.
Why here first
A single booking at a venue can be worth thousands, and venues hold lots of enquiries and years of customer history. They make money several ways, run on systems that rarely talk to each other, and make the same commercial decisions every week. Costs are rising faster than revenue, so selling more of what they already have is the cheapest growth there is.
Three jobs a venue could hand over, asked for the way an owner would ask. Pick one.
What the owner asks for
Every Monday at 8, give me every enquiry from the weekend: who’s still waiting for a reply, which look like our best bookings, and the one to ring first. Keep it to what fits on my phone.
What it reads
Enquiries inbox
Bridebook
Function diary
Built from your own inbox, Bridebook and diary. Nothing new to log into.
Monday 6 October · 08:00
11 enquiries over the weekend. 4 still waiting for a reply.
Ring first: Priya and SamSaturday 12 June 2027 · 110 guests · asked about exclusive use · arrived Friday 21:40, no reply yet
Tom and HannahViewing requested for any Saturday in October · 80 guests
A local firmChristmas party, 60 people · booked with you the last two years
A person decides who to ring. The brief only says why.
Illustration Made-up names, dates and numbers, to show the idea. What it would say for your venue is what a pilot finds out.
Six moments
The Friday night enquiry
21:40 on a Friday. A couple sends the same message to four venues, then goes to bed. On Monday morning it’s one of forty in the inbox, answered in the order it arrived. By then one of the other three has already offered a show-round.
Why it happens. Most enquiries arrive out of hours, couples contact several venues at once, and replies go out in arrival order because nothing says which ones matter most.
700 wedding enquiries in a year. 60 book. Everyone studies the 60. The lessons are in the 640: some were never a fit, some went quiet after the brochure, some chose elsewhere, some waited too long for a reply. Which group was worth another go?
Illustration A made-up example to show the idea, not a client result.
Why it happens. Lost enquiries rarely get a recorded reason, so the pattern is never seen. The records to find it usually already exist, spread across an inbox, a spreadsheet and a booking system.
What we’d do
Join two or three years of enquiries to what actually booked and find the step where good ones slip away.
What we won’t
Put a £ figure on it that the records can’t support.
Every Monday someone pulls figures from the booking system, the sales tool and a spreadsheet, so the owner can see how last week went. By the time it’s ready, the week has started.
Why it happens. The systems don’t talk, so a person becomes the integration.
Your wedding coordinator hands in their notice in March. Forty couples lose the person who knew them: the first-dance song, the aunt who can’t do stairs, the deposit that was agreed late.
Why it happens. The record holds what happened. The person held what it meant.
Staff turnover in UK hospitality has fallen from 75% to 67%: still among the highest of any sector.ClaimFourth, via The Caterer
What we’d do
Every enquiry and conversation in the shared record, and a handover pack when someone moves on.
What we won’t
Monitor individual staff. Numbers are for the team, not a league table.
A guest asks an AI assistant for a country-house hotel near the Lakes with a wedding licence. Your venue isn’t in the answer. Two chains and a booking site are.
Why it happens. AI assistants are starting to book, and the first partners are the big platforms.
Google began testing hotel booking inside its AI search in the US; independent hotels weren’t among the first partners.ClaimHotel Online, Aug 2026
What we’d do
Check it: five prompts, screenshots, the same check each month. Fixes go to a web specialist.
What we won’t
Promise to get you “ranked in ChatGPT”. Nobody can honestly promise that.
Know who to approach and when, research them properly, write to them well, and follow up.
Our working view, being testedMidweek and winter dates sit empty while Saturdays in June go three times over.Quiet periods stay quietMore
RevenueActs up to: Do with approvalOur working view, being tested
Why it happens. Demand for quiet dates has to be found, and nobody has the list of who might want them.
What the system does. The customers most likely to book quiet dates, found in your own history.
Our working view, being testedA three-star review about the breakfast, answered two weeks later, or never.Nobody treats reviews as part of sellingMore
RevenueActs up to: DraftOur working view, being tested
Why it happens. Nobody owns reviews, and there’s no time to reply well.
What the system does. Replies drafted in your own voice, and a same-day flag when something goes wrong during a stay.
Some outside evidenceA couple has chosen three venues before sending a single email.You’re not on the shortlist when customers start lookingMore
RevenueActs up to: SuggestSome outside evidence
Why it happens. The places people look (search, marketplaces, recommendations) aren’t watched, so nobody sees the gap.
What the system does. A regular check of where your customers look and whether you appear, with the gaps ranked by what they’re worth.
Our working view, being testedAsked for a Lakes wedding venue, the assistant names two chains and a booking site.AI search doesn’t mention youMore
RevenueActs up to: DraftOur working view, being tested
Why it happens. AI assistants rely on what other sites say about you, and nobody is tracking it.
What the system does. The same questions asked every month across the main assistants, what they say, and what to change. Fixes to the site go to a web specialist.
Some outside evidenceBridebook, Instagram or the wedding fair: which one actually fills Saturdays?Nobody knows which marketing brings customers who buyMore
RevenueActs up to: ShowSome outside evidence
Why it happens. Where each enquiry came from is rarely recorded, and never joined to what it was worth.
What the system does. Every enquiry traced to its source and joined to what it became, so each channel shows its cost per booking.
Our working view, being testedThe photographer who sent five couples last year has never heard from you.The people who recommend you aren’t looked afterMore
RevenueActs up to: DraftOur working view, being tested
Why it happens. Referrals arrive informally, so nobody sees who sends the most.
What the system does. Referral sources recorded, the best ones shown, and a simple routine for keeping in touch with them.
Our working view, being testedWhich local firms hold Christmas parties, conferences or away days? Nobody has the list.Nobody knows who to go afterMore
RevenueActs up to: SuggestOur working view, being tested
Why it happens. Building the list takes research nobody has time for.
What the system does. A list of the organisations that look like your best past customers, with the reason each is on it.
Some outside evidenceA firm has just opened an office ten miles away and hired forty people.Nobody knows when a prospect is readyMore
RevenueActs up to: SuggestSome outside evidence
Why it happens. The signals (a new office, a big hire, a new event manager) are public but scattered.
What the system does. Changes that suggest a prospect may need you now, checked and dated, on a weekly list.
Some outside evidenceFinding who books the Christmas party at each local firm.Researching each prospect takes too long, so it doesn’t happenMore
Time backActs up to: DraftSome outside evidence
Why it happens. The information is spread across websites, LinkedIn and news.
What the system does. A short, checked brief on each prospect: who to write to, what’s changed, and one honest reason to get in touch.
Our own data“We are a beautiful venue…” sent to two hundred firms.Outreach is generic, and gets ignoredMore
RevenueActs up to: Do with approvalOur own data
Why it happens. Writing something specific to each person takes time, so the template wins.
What the system does. Drafts written for each person from the research, with one real question. A person reads and approves every one before it’s sent.
Our own dataA season of emails to local firms, and no way to say if any booked.Nobody can tell what the outreach achievedMore
CapabilityActs up to: ShowOur own data
Why it happens. Outcomes aren’t recorded against the messages that started them.
What the system does. Every approach tracked from first message to booking, so you can see what works and stop what doesn’t.
Our working view, being testedCoach tours, golf societies and wedding planners who could fill quiet weeks.Groups and trade bookings aren’t pursuedMore
RevenueActs up to: DraftOur working view, being tested
Why it happens. They need a different kind of approach, and nobody owns it.
What the system does. A list of the groups and trade partners that fit, and a simple routine for approaching and keeping them.
ConvertTurn enquiries into bookings
Reply fast, with the right facts, a clear next step and a fair price.
Strong outside evidenceA couple emails four venues on a Friday night. The first proper reply gets the show-round.Good enquiries go cold before anyone repliesMore
RevenueActs up to: Do with approvalStrong outside evidence
Why it happens. Nothing tells the team which messages matter most, so everything waits its turn.
What the system does. Every enquiry answered quickly and honestly, and the ones worth a call first put in front of the right person.
Fact Companies that replied to a web enquiry within an hour were nearly 7 times as likely to reach the decision-maker as those that waited even one hour longer. 23% never replied at all. Harvard Business Review, 2011 (2,241 US companies)
Our working view, being testedA 120-guest Saturday wedding and “how much for a party?” wait in the same inbox.Nobody knows which opportunities to work firstMore
RevenueActs up to: SuggestOur working view, being tested
Why it happens. What predicts a sale is buried in old records nobody has time to study.
What the system does. Evidence from your own history about what tends to turn into business, tested against the simple rule you use now.
Our working view, being testedA Tuesday in January is priced like a Saturday in June.Prices are set on gut feelMore
RevenueActs up to: SuggestOur working view, being tested
Why it happens. What people paid, and what they walked away from, sits in different systems.
What the system does. Price evidence by date, product and customer type, from your own records.
Some outside evidencePrice, capacity, rooms and parking, before anyone has asked about the day itself.The same questions are answered all dayMore
Time backActs up to: Do with approvalSome outside evidence
Why it happens. The answers exist, but in people’s heads and old emails.
What the system does. Draft replies from your own facts and wording. A person checks and sends.
Our working view, being testedEvery wedding proposal is written from scratch by the one person who knows the packages.Documents start from a blank pageMore
Time backActs up to: DraftOur working view, being tested
Why it happens. Most of the content already exists, but nobody has time to assemble it.
What the system does. First drafts assembled from the record and your own templates, for a person to edit.
Our working view, being testedThe coordinator is on a show-round when the phone goes.Calls are missed and never returnedMore
Time backActs up to: DraftOur working view, being tested
Why it happens. Nobody sees a list of who called and wasn’t answered.
What the system does. Missed calls on the day’s list, with who they were if you’ve heard from them before.
Our working view, being testedA couple asks for prices, gets the PDF, and never books a show-round.Interest doesn’t turn into a visitMore
RevenueActs up to: Do with approvalOur working view, being tested
Why it happens. The next step is left to the customer, and nobody follows it up.
What the system does. A clear next step in every reply, and a gentle nudge when a visit hasn’t been booked.
Our working view, being testedJune Saturdays are fully booked; the couples who wanted one are simply told no.When you’re full, the business walks awayMore
RevenueActs up to: Do with approvalOur working view, being tested
Why it happens. Nobody records who was turned away, so nobody offers them anything else.
What the system does. Every turned-away enquiry offered the next best option, and counted.
Our working view, being testedA room booking abandoned at the payment step on a Sunday evening.Online bookings are abandoned halfwayMore
RevenueActs up to: Do with approvalOur working view, being tested
Why it happens. Nobody sees the half-finished bookings, so nobody follows them up.
What the system does. Abandoned bookings noticed, and a helpful message drafted where you’re allowed to send one.
Our working view, being testedA coach-tour enquiry sits in reception’s inbox for a week.Group and event enquiries get lost in the general inboxMore
RevenueActs up to: Do with approvalOur working view, being tested
Why it happens. The inbox treats every message the same.
What the system does. Group and event enquiries recognised as they arrive and sent straight to the person who handles them.
RecoverRescue what’s slipping away
Quotes nobody chased, deals gone quiet, and lost business nobody asked about.
Our working view, being testedThe brochure goes out on Monday. Nobody hears back, and nobody asks.Follow-up stops after one or two triesMore
RevenueActs up to: Do with approvalOur working view, being tested
Why it happens. Chasing is nobody’s job, and it’s the first thing to go on a busy day.
What the system does. A follow-up routine that runs itself, with a person deciding what gets sent.
Our working view, being testedWas it the price, the date, or the three days it took to reply? Nobody knows.Nobody records why business was lostMore
RevenueActs up to: Do within agreed limitsOur working view, being tested
Why it happens. Recording a loss feels like admin, and nobody looks at it afterwards.
What the system does. A one-click reason on every loss, and a monthly look at the pattern.
Our working view, being testedA deposit is three weeks late, and nobody notices until the balance is due.Money leaks where nobody’s lookingMore
RevenueActs up to: Do within agreed limitsOur working view, being tested
Why it happens. Each one is small, and nobody owns the gap between systems.
What the system does. Payments matched to what’s due, and a weekly list of what needs chasing.
Some outside evidenceA couple has been “thinking about it” for six weeks since the show-round.Deals stall with “we’re thinking about it”More
RevenueActs up to: DraftSome outside evidence
Why it happens. There’s no clear next step and no reason to decide now.
What the system does. The stalled ones picked out with what usually moves people like them, and a drafted, honest nudge.
ReactivateBring past customers back
Former customers, lapsed regulars and old enquiries who would come back if someone asked.
Our working view, being testedA firm has had its Christmas party with you three years running. It’s October and they haven’t called.Past customers are never asked backMore
RevenueActs up to: Do with approvalOur working view, being tested
Why it happens. Repeat business is rarely anyone’s job, so its absence shows up late.
What the system does. A flag when a regular is later than usual, with what they bought before and who to call.
Some outside evidenceBooking sites take 15 to 25% of room revenue on the bookings they bring.Too much of the revenue goes to middlemenMore
RevenueActs up to: SuggestSome outside evidence
Why it happens. Going direct takes effort, and past customers are hard to reach again.
What the system does. The direct share tracked, and past customers brought back where you’re allowed to contact them.
Some outside evidenceEighty wedding guests, some of them next year’s couples.Guests are never seen as future customersMore
RevenueActs up to: DraftSome outside evidence
Why it happens. Only the person who booked is ever recorded.
What the system does. Where guests agree to it, a way to stay in touch, and a gentle invitation when the time is right.
Our working view, being testedA couple married here five years ago, never invited back for their anniversary.Anniversaries and milestones pass unmarkedMore
RevenueActs up to: Do with approvalOur working view, being tested
Why it happens. The dates are on file, but nobody looks.
What the system does. A reminder when a meaningful date comes round, with a drafted message for approval.
Our working view, being testedChristmas vouchers expiring unused every spring.Gift vouchers are sold and never turned into regularsMore
RevenueActs up to: Do with approvalOur working view, being tested
Why it happens. Nobody follows the voucher to the person who uses it.
What the system does. Voucher holders reminded before expiry, and welcomed back after their visit.
ExpandGrow what customers spend
The extras, upgrades and bigger bookings they would happily say yes to.
Our working view, being testedWedding guests who’d book a room, if anyone asked them in time.Customers aren’t offered what they’d happily buyMore
RevenueActs up to: Do with approvalOur working view, being tested
Why it happens. Nobody has the time to match the right offer to the right customer.
What the system does. The right offer to the right customer at the right moment, from what similar customers bought.
Some outside evidenceA local firm has doubled in size and still books one Christmas party.Customers who could buy more aren’t spottedMore
RevenueActs up to: SuggestSome outside evidence
Why it happens. Nobody puts what they buy next to what’s changing in their business.
What the system does. Accounts showing signs they could buy more, with the evidence and a suggested approach.
Our working view, being testedEighty wedding guests, half of them staying in the town down the road.Event guests aren’t offered a room in timeMore
RevenueActs up to: Do with approvalOur working view, being tested
Why it happens. The guest list sits with the organiser, and rooms are offered too late or not at all.
What the system does. A timely, personal offer to guests where the organiser agrees, and a count of rooms filled.
RetainKeep the customers you’ve won
Spot the bookings and accounts at risk, and look after people before they drift.
Our working view, being testedAn allergy noted at booking never reaches the kitchen’s function sheet.Small mistakes reach the customerMore
Time backActs up to: Do within agreed limitsOur working view, being tested
Why it happens. Every copy between systems is a chance for an error, and nobody checks the copies.
What the system does. Details flow from the record, and changes are flagged to the people who need to know.
Some outside evidenceA couple books in March and hears nothing until the final-numbers form in August.Customers hear nothing after they’ve bookedMore
RevenueActs up to: Do with approvalSome outside evidence
Why it happens. Updates are sent when someone remembers, not on a plan.
What the system does. A simple, personal plan of updates from booking to the day, drafted for approval.
Our working view, being testedA cold room mentioned on TripAdvisor two days later.Problems are heard about in the review, not while there’s time to fix themMore
RevenueActs up to: SuggestOur working view, being tested
Why it happens. Nobody sees the early signs during the stay.
What the system does. A same-day flag when something looks wrong, so a person can put it right before the guest leaves.
Some outside evidenceA couple stops replying, misses a payment, then cancels three months out.Bookings at risk aren’t spotted until they cancelMore
RevenueActs up to: SuggestSome outside evidence
Why it happens. Nobody watches the pattern across a booking’s history.
What the system does. An early warning on bookings that look at risk, and a nudge to get in touch while it can still be saved.
Start with eight weeks
Two weeks to join up the records and learn how the team works. On day 10 you see where revenue is going missing and what it’s worth; if that doesn’t justify going on, we say so and stop. Four weeks to build the system around the biggest opportunity, with the team. Two weeks to measure it honestly.
It might solve some of it. That’s one of the first things I’d look at. I’m not trying to replace technology that already works and is already used. But often the information sits across several systems, or the CRM records things without turning them into the decisions and actions the team needs. I only build where there’s a genuine gap.
Is this just ChatGPT?
No. AI is a tool inside the system, not the product. I use the simplest reliable method for each problem: Python or SQL for calculations and cleaning data, statistical models, machine learning where there’s enough evidence, and AI models where they’re strongest, reading notes and emails, reasoning and explaining. The business problem first, how people use it second, the technology third.
What if the numbers don’t justify a build?
At Day 10 I share a short report: where revenue is being missed, the evidence behind it, what it could be worth where I can estimate it responsibly, and what the system would do first. If the opportunity isn’t big enough to justify the build, I’d rather tell you than manufacture work. You keep the report and we stop there.
What would you need from us?
One person who owns the work on your side, access to the right people and records (exports from your booking system, sales tool and enquiry inbox), and a team willing to try what we build and tell me what they think. It can’t create value if nobody uses it, so it has to be collaborative.
How will we know if it worked?
Before anything changes, we agree a baseline and what success means. Then we measure it: reply times, show-rounds, bookings, hours saved, and whether the team actually uses it. I separate opportunity found from value actually realised, and I don’t claim revenue I can’t attribute to the work.
What happens to our data?
Your data is yours, always. It’s held encrypted, used only for this work under a written data processing agreement, and deleted at the end with a certificate unless you ask me to keep it. If we stop, you’re not trapped. I’d rather be useful enough that you don’t want to leave than make leaving hard.
You’re young, and there’s no hotel case study. Why take the risk?
Fair question. I’m 18, and there’s no hotel case study yet, which is why the first clients are founding partners and why the pilot has a Day 10 stop. I’m not asking you to believe claims I haven’t earned. What I have done is the preparation: researched the problems, built and tested the methods on a made-up hotel, and thought carefully about security. The pilot is how we find out whether that turns into real value for you.
Why hotels and venues?
They have valuable bookings, lots of enquiries and years of customer history, several ways of making money, systems that don’t talk to each other, and people making the same commercial decisions every week. It’s where I’m starting, not the only place this works.
Do you do websites, marketing or social media?
No. I’m not trying to be your marketing agency, CRM supplier or IT support. I focus on problems where your own records can improve decisions and actions. If a specialist product or provider can solve something better, I’d rather recommend them than build a worse version.
Who will I work with?
Me. I lead the work personally, with AI tools I’ve built and tested, and specialists only where they’re needed and agreed with you.