Worked scenario · Hotels and restaurants
Three venues, three tools, no overall view.
A family group: a 48-room hotel and two restaurants, 60 staff, three systems that don't talk to each other, and booking requests arriving seven days a week, often outside opening hours. Here is a full engagement.
The company
- Size
- 60 staff across three venues, under one management team
- Business
- A 48-room hotel and two restaurants, leisure and business guests
- Volume
- Around 1,800 stays and 45,000 covers a year
- Tools in place
- A PMS for the hotel, one till system per restaurant, two distribution platforms, a rota spreadsheet
- What is missing
- No consolidation across the three venues. The same guest exists twice.
The starting point
On Monday morning, management reassembles the weekend by hand: occupancy in the PMS, covers and average spend in two different till systems, platform bookings in a third interface. An hour and a half every week, for a picture that is already out of date.
The simplest questions have no reliable answer. Did the guest who stayed three nights eat at the restaurant? What share of revenue comes from the platforms once their commission is deducted? Is the first restaurant's average spend really higher than the second's, or do the two tills simply not count the same thing?
Meanwhile, booking requests arrive by phone, by email and through social messaging, often in the evening and at weekends. Those arriving outside service hours wait until the next day. Some never call back.
What the audit puts on the table
- 01
The same guest exists twice, once in the PMS and once in the tills, with no shared identifier. There is no way today to know that a hotel guest ate at the restaurant.
- 02
The two tills do not compute average spend the same way: one includes complimentary drinks, the other does not. Any comparison between the two venues is wrong until that is settled.
- 03
Platform commissions appear nowhere in the figures management tracks. Revenue is shown gross, and the real margin per channel is unknown.
- 04
The PMS exposes a decent API, so do the tills. Distribution goes through regular exports, sufficient for a daily rhythm.
- 05
Over a month, a significant share of booking requests received outside opening hours never got an answer. That volume was measured nowhere.
What gets built, week by week
- Weeks 1-2
Connectors and ingestion
The PMS, both tills, and the platform exports. Daily, incremental ingestion with retry on failure. The rota spreadsheet is taken over once, then fed from the system.
- Weeks 3-4
Warehouse, model and guest matching
A model that names the business: guest, stay, cover, venue, booking channel. Guests present in both the PMS and the tills are matched on verifiable attributes, never on an approximation. Anything that stays doubtful is flagged as such rather than merged wrongly.
- Week 5
The definitions, settled
One session with management and the two floor managers: what a cover is, what average spend is, how complimentary items are treated, and whether tracked revenue is gross or net of commission. Each rule is written down and implemented once.
- Weeks 6-7
Group dashboards and alerts
Occupancy, RevPAR, covers, average spend, channel split with commissions deducted, and a comparison across the three venues on a finally identical basis. Alerts on occupancy for an upcoming weekend, on a channel slipping, on an unusual swing in average spend.
- Weeks 8-10
Booking on direct channels
An agent handles requests arriving on your direct channels from your real availability and your rules, not from its imagination. It proposes, confirms what can be confirmed, and hands over as soon as there is an exception: a group, a special request, an allergy, an event. The request that arrives at 10:40 pm gets an answer at 10:40 pm.
What it changes
Reassembling the weekend
About 1 h 30 by hand, every Monday
Available every morning, with no intervention
Group view
Three tools, three partial pictures
Three venues on a comparable basis
Revenue by channel
Gross, commissions ignored
Net of commission, by channel and by venue
Requests received outside hours
Handled the next day, sometimes never
Immediate answer on direct channels, escalation for exceptions
Hotel guest identified at the restaurant
Impossible
Matched where the attributes allow it, flagged uncertain otherwise
These orders of magnitude are what this kind of engagement produces when the conditions below are met. They are not contractual commitments: the pricing for your own situation comes out of the audit.
What it costs, and how long it takes
Audit: €2,500, one to two weeks. Data foundation and dashboards: €16,000, five weeks. Booking agent on direct channels: €14,000, four weeks that partly overlap the previous ones. Around €32,500 over ten weeks, decided in three steps. A monthly operating fee applies. You own the code and the credentials.
What would make this scenario fail at your company
A PMS with no API
Some older versions and some entry-level plans expose nothing. Workarounds exist: they cost more and are more fragile. The audit checks this first.
Every venue with its own vocabulary
If the two restaurants refuse to align their definition of a cover and of average spend, the dashboards will stay individually correct and mutually incomparable. That is not a technical question.
Front of house kept out of it
A booking agent answering on the team's behalf without the team having defined its rules will make commitments the house cannot keep. The people who welcome guests must write the scope, not inherit it.
Forcing guest matching
Merging two guest records on a mere name similarity creates lasting errors and a GDPR risk. A partial, honest match beats a database that only looks clean.
The building blocks involved
The data audit
Two weeks at most, a fixed price, and a document that belongs to you. It is the only honest way to know what your data allows before committing a budget to it.
The data foundation
Four to eight weeks for your data to live in one place, historised and reliable. Delivered in usable stages, not in one go six months out.
The AI assistant
An assistant that answers your business questions in plain language, on figures that come out of your own database. It arrives once the foundation is reliable, never before.
The other scenarios
B2B trade and distribution
Monthly revenue lands twelve days after close, stockouts are discovered the day the customer calls, and four tools give four different definitions of revenue.
Manufacturing and after-sales
Four people in after-sales spend their days answering the same six questions: lead time, order status, part compatibility, warranty, documentation, scheduling a visit.
Do your venues speak the same language?
Thirty minutes are enough to know what your tools actually expose and where to start. If a simple export is all you need, we will tell you.