A hotel with five food and beverage outlets usually has five guest databases, five sets of regulars and no idea how much overlap there is between them. The guest who has breakfast in the café four times a week and celebrates a birthday at the rooftop bar twice a year is two unrelated records, and nobody in the building knows they are the same person.
Fixing that is the single highest value thing a multi outlet property can do with its guest data. It is also harder than it looks, because the operational questions arrive before the technical ones and most vendors will not raise them until you are already live.
This is the sequence of decisions to make, in the order they matter.
Decision one: one currency or several
The first fork. Do points earned in the café spend at the rooftop bar?
One shared currency is the version guests expect and the version that produces the behaviour you want, which is cross outlet trial. It is also the version that creates internal friction, because the rooftop bar now discounts food that the café sold. If your outlets are separately profit and loss accountable, this is a real conversation and you should have it before launch rather than in month three when the first outlet manager notices.
Separate currencies per outlet avoids the internal transfer problem entirely and delivers a fraction of the value. You have effectively run five small loyalty programmes and paid for one platform.
A hybrid works well in practice. Points are earned anywhere and redeemed anywhere, but the internal accounting recognises a transfer between outlets at cost rather than at menu price. The guest sees one simple programme, and finance sees a defensible allocation.
Whichever you pick, decide it before you configure anything. Retrofitting a currency decision after members hold balances is genuinely painful and visible to guests.
Decision two: how the internal transfer is settled
If you chose a shared currency, someone has to own the redemption cost.
Three workable approaches:
- Cost transfer. The redeeming outlet bills the earning outlet at food cost rather than menu price. Fairest treatment, most accounting overhead.
- Central marketing budget. Redemptions are absorbed by a property level marketing line rather than by either outlet. Cleanest operationally, and it keeps outlet managers from resisting redemptions, which is the failure you are trying to avoid. It requires a property level owner with a budget.
- No transfer. The redeeming outlet absorbs it. Simple, and it creates exactly the wrong incentive. Outlet managers who bear the cost of other outlets’ point issuance will find reasons why redemption is inconvenient today.
The central budget option is usually the right one for a single property. It reflects the truth of the situation, which is that the programme is a property level investment in guest retention rather than an outlet level promotion.
Decision three: earn rates per outlet
A flat rate across all outlets is simplest and rarely correct. Your café has a different margin profile from your rooftop bar, and a flat percentage means the low margin outlet is subsidising the high margin one.
Set earn rates against margin rather than against revenue, and remember this is a pricing decision. If you award one point per dirham and redeem at one hundred points for a twenty dirham reward, you have set a two percent discount on every qualifying transaction in that outlet. Whoever owns margin should sign that off. We cover the liability arithmetic in more detail in our breakdown of what a loyalty programme actually costs.
Also decide what does not earn. Service charge, tourism fees, delivery platform orders, banqueting and anything already discounted are common exclusions. Write the list down before launch, because the first time a guest disputes it you will want a policy rather than an improvisation.
Decision four: the guest identity problem
This is where the technical work sits.
To run one profile across five outlets you need a reliable way to recognise the same person at each of them. Mobile number is the practical primary key in this market. Email is unreliable because guests give different addresses, and name matching produces false merges that are worse than duplicates.
Three requirements follow.
A single enrolment path. If each outlet enrols guests through a different mechanism, you will generate duplicates on day one. One QR, one form, one record.
Deduplication on entry. The system should detect an existing member at the point of enrolment rather than creating a second record and requiring a manual merge later.
A merge tool. Duplicates will happen anyway. Your staff console needs a way to combine two records without losing visit history or points balance, and it should be restricted to a permission level above floor staff.
Test the merge behaviour during your evaluation. Ask the vendor to create a duplicate deliberately and merge it in front of you. This is not a feature anyone demonstrates unprompted and it is the one you will use most in year one.
Decision five: per outlet configuration that must stay separate
Sharing a guest profile does not mean sharing everything. These need to remain outlet specific:
- Opening hours and service periods, including any outlet that runs past midnight
- Table capacity and booking rules
- Staff permissions, so the café host cannot amend rooftop reservations
- Menus and reward catalogues, since a free coffee makes no sense at the bar
- Booking confirmation and reminder message content
The overnight service case deserves particular attention. An outlet open from eight in the evening until two in the morning is a common configuration in Dubai and it breaks a surprising number of booking systems, because a naive implementation treats a closing time earlier than an opening time as invalid. Ask any vendor to demonstrate a two in the morning booking slot. It is a fast and revealing test.
Decision six: what the redemption moment looks like on the floor
Everything above is theory until a guest tries to redeem at a till during a busy Friday service.
The mechanics that survive real service:
QR to earn, QR to redeem. No app download, no plastic card, no member number recited across a counter. Enrolment and redemption both need to fit inside the time a card payment takes.
Staff confirmation on the redeeming side. The member initiates, the staff member confirms in the console. This prevents accidental redemptions and gives you an audit trail per outlet.
Offline tolerance. Rooftop and poolside outlets have unreliable connectivity. Ask what happens when the tablet drops off the network mid transaction.
Under one minute of training. If the redemption flow needs a laminated card by the till, it will not survive staff turnover.
Decision seven: reporting that answers the group question
Per outlet reporting is table stakes. The reports that justify the programme are the cross outlet ones:
- How many members visit more than one outlet, and how that number moves over time
- Revenue per member versus revenue per non member
- Points issued by outlet against points redeemed by outlet, which is your internal transfer picture
- Outstanding liability across the property
- Which outlet acquires members who then spend elsewhere, which is usually the café and usually undervalued
That last metric changes internal conversations. The outlet enrolling the most members often has the lowest average spend, and without cross outlet attribution it looks like the weakest performer while actually feeding the whole property.
Decision eight: the message channel
Once you have one profile across five outlets, you can segment properly. A guest who visits the café weekly but has never been to the rooftop is an obvious campaign audience.
In this market that message needs to go by WhatsApp. Email open rates in food and beverage sit in the teens. Use the official WhatsApp Business Platform rather than an unofficial gateway, capture opt in explicitly at enrolment, and be aware that template approval carries a lead time you need to plan around. Consent capture is also a compliance requirement under UAE data protection law, not just good practice.
Our marketing automation service covers the segmentation and journey side of this if you want the campaign layer as well as the data layer.
A realistic sequence
Do not launch all five outlets simultaneously.
Start with two, ideally one high frequency outlet like a café and one destination outlet like a restaurant or bar. That pairing tests the cross outlet mechanic immediately, which is the whole point, while keeping the operational surface small enough to fix problems quickly.
Run it for four to six weeks. Watch enrolment rate per shift, redemption rate, duplicate creation and staff compliance. Fix what breaks. Then add the remaining outlets, by which point you have a trained team who can help onboard their colleagues.
The part that decides success
None of this is really a technology problem. The platform question is genuinely the easy part.
What determines whether a multi outlet programme works is whether the outlet managers believe it helps them. If the rooftop manager thinks she is absorbing the café’s discounts, redemption will quietly become difficult at the rooftop. Settle the internal transfer question first, put the cost somewhere neutral, and give every outlet manager a report that shows what the programme brought them rather than what it cost them.
Get that right and the technical configuration follows easily. Get it wrong and the best platform available will sit unused behind a host stand.
If you are working through this for a property in the UAE, we build exactly this and we are based in Dubai. Send us your outlet list and we will tell you where the friction will be.
