One source of truth for the whole group.
Every system you already run — property management, point of sale, accounting, guest reviews — reconciled into one set of numbers, reported to the standard your auditors and lenders recognise, and delivered to the people who decide.
Ask four people in a hotel group what last month's average rate was and you will get four answers, each defensible. The property system counts one way, the revenue manager another, finance a third, and the board sees whichever number reached the deck. Score settles it — the figure is defined once, calculated once, and everything else reads from there.
Score has two halves, and most reporting products only have the first. The analytical half is the warehouse and the dashboards: six families of screens covering revenue, demand, restaurants, spa, bookkeeping and management accounts, with drill-down from the group to a single room night.
The operational half is delivery. A general manager does not open a dashboard every morning; the daily summary reaches them where they already are, three times a day. Purchasing does not open the accounting system daily; the payment list arrives by email. And when a source goes quiet, a watchdog says so — because a stale number on a screen looks exactly like a fresh one.
Together they close a loop: you see the trend, you act, and the next day's figures tell you whether the action landed — on numbers you did not have to assemble by hand.
The reporting, screen by screen
Six of the screens, with sample numbers in place of anyone’s real ones. The last two are the ones almost nobody else ships: whether today’s figures can be trusted, and who in your group is allowed to see them.
Revenue by month — this year against last
Where the bookings came from
By property — RevPAR, month to date
| Property | RevPAR | Against budget | Revenue | |
|---|---|---|---|---|
| Harbour House | €148 | ▲ 96% | €1.24m | |
| Old Mill | €136 | ▲ 88% | €0.91m | |
| Cliffside | €129 | ▲ 84% | €1.06m | |
| Riverside | €121 | ▼ 78% | €0.74m | |
| Garden Court | €118 | ▲ 76% | €0.68m | |
| The Orangery | €104 | ▼ 67% | €0.52m |
Occupancy by month and day of week
How early the rooms sell
Length of stay
Full houses and what they cost you
| Nights sold out this year | 34 |
| Of those, on a weekend | 27 |
| Requests turned away | 612 |
| Average rate on those nights | €214 |
| Rate headroom, estimated | €9–14 |
Revenue by outlet
Covers by hour
Menu engineering — popularity against margin
| Item | Sold | Share of covers | Category |
|---|---|---|---|
| Grilled sea bass | 1,284 | Star | |
| Lamb shoulder | 1,061 | Star | |
| Garden risotto | 742 | Workhorse | |
| Local cheese board | 418 | Puzzle | |
| Duck confit | 196 | Dog |
From revenue to gross operating profit
Budget against actual, by property
| Property | Actual | Budget | Variance |
|---|---|---|---|
| Harbour House | €412k | €388k | +6.2% |
| Old Mill | €301k | €295k | +2.0% |
| Cliffside | €356k | €361k | −1.4% |
| Riverside | €248k | €262k | −5.3% |
| Garden Court | €227k | €218k | +4.1% |
| The Orangery | €174k | €186k | −6.5% |
Where the operating cost sits
The screen nobody else ships. It answers the only question that matters before a board meeting: can I rely on what I am about to present?
Source freshness
| Source | Last delivered | Rows | State |
|---|---|---|---|
| Property management | today 03:12 | 14,208 | On time |
| Point of sale | today 03:20 | 9,671 | On time |
| Spa system | today 03:21 | 412 | On time |
| Accounting ledger | today 03:44 | 31,006 | On time |
| Guest reviews | yesterday 21:05 | 188 | Late, expected |
| Parking | 3 days ago | — | Not delivering |
Completeness, last fourteen days
| Days complete | 12 of 14 |
| Days reported as partial | 2 |
| Days silently wrong | 0 |
| Rows held in quarantine | 37 |
| Corrections reloaded | 4 |
In a group, the question is never only what the reporting shows — it is who is allowed to see it. A general manager should see their own hotel in full and nobody else’s payroll. An owner should see everything. A lender should see one schedule and nothing more.
Who sees what
| Screen | Owner | General manager | Financial controller | Revenue manager | Observer | Administrator |
|---|---|---|---|---|---|---|
| Group overview | ● | ◐ | ● | ● | ● | ● |
| Revenue and demand | ● | ◐ | ● | ● | · | ● |
| Restaurants and spa | ● | ◐ | ● | · | · | ● |
| Bookkeeping | ● | · | ● | · | · | ● |
| Management accounts | ● | ◐ | ● | · | · | ● |
| Supplier payments | ● | · | ● | · | · | ● |
| Data health | ● | ◐ | ● | ● | · | ● |
| People and access | · | · | · | · | · | ● |
Access log
| Who | Role | Last seen | State |
|---|---|---|---|
| A. Bellweather | Owner | today 08:04 | Active |
| M. Sorensen | General manager | today 07:41 | Active |
| J. Aldridge | Financial controller | yesterday 18:22 | Active |
| P. Novak | Revenue manager | today 09:15 | Active |
| Auditor, seasonal | Observer | 11 days ago | Suspended |
The rule underneath it
Empty means nothing, not everything
If nobody has granted a person a screen, they see no screens. Forgetting to set access has to look like no access — never like full access.
A seat, not a shared login
Each person signs in as themselves, so the access log is a record of people rather than of a password that six of them know.
Scoped to the property
A manager of one hotel sees that hotel. The group roll-up exists for the people whose job is the group.
Read-only into your systems
Score connects to source systems with read-only rights. It reports on your data; it cannot write to your property system or your ledger.
Six families of screens
Revenue and commerce
Where did the money come from, and what did it cost to get it?
- Rate, occupancy and RevPAR by property
- Segments, channels and commission
- Ancillary revenue and discounting
- Guest origin, new against returning
Demand and occupancy
How full are we, how early do people book, and where is the room left to sell?
- Occupancy by property and season
- Booking window and pace
- Length of stay distribution
- Sell-out days and rooms out of order
Restaurants and bars
Which outlets earn, which dishes carry them, and how many hotel guests ever eat in?
- Outlet ranking and covers by hour
- Average check and items per guest
- Menu engineering and slow movers
- Capture rate from rooms to outlets
Spa and wellness
Is the spa a service or a business, and how much of the house ever walks in?
- Revenue, treatments and average spend
- Service mix and retail against treatments
- Guest penetration from rooms
- Seasonality and repeat visits
Bookkeeping
What do the books say, in the form your accountant already signs?
- Account movements and trial balance
- Cash flow by category, bank and till
- Payroll, taxes and contributions
- Receivables and payables by age
Management accounts
What is actually left after the hotel has been run — in a standard a lender recognises?
- Profit and loss to the USALI layout
- Cost per occupied room and flow-through
- Departmental result by property
- Budget against actual, month and year to date
Why the numbers hold
Reporting does not usually fail by crashing. It fails by quietly disagreeing with itself until nobody trusts it any more. These are the rules we build in so that does not happen — each one learned the expensive way.
Every figure comes from the source that owns it
Money comes from the financial ledger. Occupancy and rate come from the night audit. You cannot derive one from the other, and systems that try will be wrong in a way nobody notices for months.
One figure, one precision, everywhere
The same measure appears on a chart, a tile, a table and in an emailed summary. If it is rounded differently in any of them, a reader sees a discrepancy rather than a rounding choice — and stops believing the rest of the screen.
One filter boundary per screen
If the summary tiles cut the data at today and the breakdown below them does not, the screen contradicts itself even though every card is individually right. The boundary is set once, in the model, and no card may override it.
Compare only on a base that has not moved
Year-on-year comparison is meaningless if the definition changed in between. When something moves between revenue lines, the prior year has to be restated on the new basis before the two are put side by side.
Currencies are never quietly added together
A group operating across borders holds payables and revenue in several currencies. Score reports each in its own, and converts only where you have told it which rate to use and on what date.
The screen never promises what it cannot do
If a caption says click to drill in, clicking has to drill in. Interface copy that describes behaviour which was planned but never built is the fastest way to lose a user, because it teaches them the whole thing may be decorative.
You see it before you open anything
A general manager checks their phone dozens of times a day and a reporting portal perhaps once a week. So the summary goes to the phone. It runs three times a day, and it follows one rule that sounds small and is not: the first run of the day always sends, even when the data is incomplete, because the morning meeting needs a signal. Later runs only send again if the picture actually improved.
- Occupancy last night74.2%
- Average rate€178
- Rooms revenue€41,300
- Food and beverageawaiting
- Month to date against budget96.4%
- Occupancy last night74.2%
- Average rate€178
- Rooms revenue€41,300
- Food and beverage€12,850
- Month to date against budget98.1%
The payment list that stops the phone call
Supplier due dates sit in the accounting system, and nobody opens the accounting system to browse. So the list comes out by email on a schedule: due today, due this week, already overdue — each in its own currency, never summed into a false total.
A watchdog for the failures that make no noise
When a source stops delivering, yesterday's figure sits on the screen looking like today's. Nothing has crashed, so nothing alerts. A separate check watches how fresh each source is, records incidents, and raises one alert per problem rather than one per hour — because an alarm that cries daily stops being read.
It has to read the estate you already have
This is where most reporting projects quietly die. A group does not run one property system and one till. It runs whatever each hotel arrived with — different vendors, and often different versions of the same vendor at neighbouring properties. And the vendor may not give you an interface at all. None of that is a reason to abandon the project, and Score is built on that assumption rather than against it.
What we connect to
| System | What it gives Score |
|---|---|
| Property management system | Room sales, rates, segments, channels, stay length, guest origin, night audit statistics |
| Point of sale in restaurants and bars | Checks and line items by outlet, shift business date, covers, discounts, gross and net kept apart |
| Spa and leisure system | Treatments, retail, average spend, guest penetration from the rooms side |
| Accounting and management accounting | Ledger movements, cash flow, payroll, taxes, supplier payables, USALI mapping |
| Guest review platform | Review scores, full review text, category praise and complaint, competitive set |
| Access, ticketing and rental systems | Passes, rentals and takings for resorts with activity revenue |
| Parking | Entries and parking revenue as an ancillary line |
| Budget and planning workbooks | The plan side of budget against actual — which no operational system will ever hold |
And how we take the data
Integration is not the same thing as an interface. All of the following are running in production today, side by side — including two different channels to the same vendor, because the documented interface turned out not to be the one that worked.
| Method | When it is the right one |
|---|---|
| Direct read from the source database | The fullest access, when your vendor or your own team will grant it |
| The vendor's web interface | The standard route where a documented, working interface exists |
| Parsing the vendor's own portal exports | When the interface is incomplete, undocumented or simply not offered to you |
| Scheduled file drops onto a shared folder | The route that unlocks estates where nobody will ever open the database — and the one most often forgotten |
| Older enterprise service interfaces | Access, ticketing and legacy systems that predate modern interfaces and are not going anywhere |
| Structured feeds from the accounting system | Repeated intraday collection for payables and cash, not once a night |
| A managed intake folder for spreadsheets | Budgets and reference data that live in a workbook on a controller's desk, loaded without anyone retyping them |
| The night audit statistical report | Occupancy and room statistics, which financial data cannot produce correctly |
Who is allowed to see it
In a single hotel this barely matters. In a group, it decides whether the reporting can be opened up at all — because the moment one general manager can see another property’s payroll, the safe answer becomes to show nobody anything.
Roles that bend
Owner, general manager, financial controller, revenue manager, observer, administrator. The role establishes a starting set of screens, and any individual can be given or denied one without inventing a new role for them. Each person also lands on the screen their job actually starts with, rather than on a generic home page.
Nothing granted means nothing seen
If nobody has assigned a person any screens, they see none. Forgetting to set someone up has to default to no access, not full access — the opposite default is exactly how reporting leaks.
Scoped to the property
A manager responsible for one hotel sees that hotel. The group roll-up is there for the people whose job is the group. A seasonal auditor can be given one schedule and suspended again when the engagement ends.
Named people, not a shared password
Everyone signs in as themselves, so the record of who opened what is a record of people rather than of a password six of them know. That sign-in journal is itself visible only to administrators.
Read-only into your systems
Score connects to your property system, your tills and your ledger with read rights only. It reports on your data and cannot write back to the systems it reads. Credentials are held outside the codebase, and the reporting is reached over your own network rather than published to the open internet.
Recoverable, including on a bad day
The warehouse is backed up on a schedule, and the restore procedure is written down and rehearsed rather than assumed. A local copy of the components is kept beside it, so rebuilding does not depend on the internet being available at the moment you need it.
What we will not pretend
Some screens need data that a hotel may simply not keep. We would rather you knew which ones before the project than discover them at the demonstration.
| What you want | What it needs from you | What you get without it |
|---|---|---|
| Treatment room utilisation | A register of rooms and therapist schedules | Spa revenue and service mix, but not utilisation |
| Departmental profit by property | Agreed rules for allocating shared costs | A consolidated result, not one split by outlet |
| Budget against actual | A budget, in any workbook format you already use | Actuals and year-on-year comparison only |
| Food cost and menu margin | Dish costings maintained in the till or elsewhere | Popularity and revenue, without the margin side |
| Receivables and payables by age | Period-end balances, not only movements | Flows through the accounts, without ageing |
| Forward occupancy forecast | Reservation data on the books, not just arrivals | Booking pace from what has been captured so far |
Score, or build it yourself
Plenty of groups can buy a reporting tool and point it at a database. The tool is rarely the hard part. What follows is.
| The question | With Score | Building it in house |
|---|---|---|
| Who decides what a measure means | Defined once in the model, with the definition written down and agreed | Each report author decides again, and the definitions drift apart |
| A new outlet or till code appears | Shows up unmapped in a coverage report, fixed by adding a row | Falls into an unattributed bucket nobody notices for a quarter |
| A source stops delivering on a Saturday | Freshness check raises it, incident recorded, one alert not forty | The dashboard keeps showing the last good day, looking healthy |
| Accounting posts a correction after the fact | Reload is idempotent — the corrected day replaces itself, without doubling | Corrections either double-count or are silently dropped |
| A property migrates to a different system | Authority is recorded by date, so the history stays continuous | The series breaks at the migration and comparison stops working |
| Someone has to run it every day | Scheduled loads, load journal, quarantine for bad rows, backups and restore | An analyst who becomes the single point of failure and then leaves |
Cadence sets the price, Baton runs the day, Melody serves the guest, Refrain keeps them coming back — and Score proves what worked, so next month's price is set on evidence. See the full platform.
Questions owners actually ask
Our property system is not a big international brand. Does that rule us out?
No. What matters is not the name of the system but whether it can hand over data in any of the ways listed above — a database connection, an interface, a scheduled export, even a file dropped on a folder. Smaller and regional systems are frequently easier here, not harder, because their exports are simpler.
Our vendor will not give us access to anything. Now what?
Then we work from what the system already produces for you: scheduled exports, the night audit statistical report, portal downloads. This is a common situation in managed and branded properties and it is not a blocker. It changes how often data arrives, not whether the reporting works.
Our hotels run different systems, and two of them run different versions of the same one.
That is the normal case and it is already running in production. Each property is connected on its own terms, and the group roll-up is built on the harmonised model rather than on the assumption that every hotel looks alike.
How often do the numbers refresh?
Loads run to a schedule set around when each of your sources actually produces data — there is no point collecting at six if the export lands at ten. The daily summary goes out three times a day. Where a source supports more frequent collection, such as supplier payables, it is collected several times a day.
What happens when accounting corrects a figure for a month we already reported?
The affected period is reloaded rather than patched, so the corrected version replaces the old one cleanly and nothing is counted twice. The load journal records that it happened, which matters when someone asks months later why a number changed.
How will I know the report is wrong before my board does?
Three ways. Rows that fail validation go to a quarantine table instead of being silently zeroed. Freshness of every source is checked independently of the loads themselves. And the daily summary carries its own completeness marker, so a partial day is labelled as one rather than presented as final.
Can we get our own measures, not just the ones you ship?
Yes. Measures are defined in the model layer rather than inside individual screens, which is what keeps them consistent everywhere they appear. A new measure is added there once and becomes available across the reporting.
We operate in more than one currency.
Each amount is reported in the currency it was transacted in. Nothing is converted or combined unless you have specified the rate and the date to use, because a total that silently mixes currencies is worse than no total at all.
Who receives the daily summary, and can we change the list ourselves?
Recipients are managed per distribution, and the daily summary and the payment list are kept deliberately separate — different lists, different schedules, different checks. One is never allowed to mark the other as delivered.
Does Score replace our accounting system?
No, and it should not. Your accounts remain the authority for what was posted. Score reads from them, reconciles them against the operational systems, and reports the result in the layout a lender or a buyer expects.
Do we have to take the rest of the YMME platform to use Score?
No. Score is built to sit on top of whatever you already run, and that is the point of it. It works alongside our own products and it works with systems that are nothing to do with us — which is the same principle the whole platform is built on.