Five situations.
Five individual perspectives at the table.

The market leader bought personalisation in 2019 and never gave it back. An industry rarely gets a clearer signal. Personal menus are an operating system, not a gimmick. Only the order of the followers is open.

Personalisation needs no new app, only a layer behind it: order history, rules, delivery. The existing app stays and gets connected. That is an integration project with a manageable scope.

Time of day, location, history: the segments in food service are simpler than in retail and work faster. A breakfast guest is a breakfast guest. That makes QSR the most rewarding field for data-driven offers.

The business case runs through the ticket. Recommendations and fitting bundles lift the basket without a cent of discount. That sets personalisation apart from almost every other marketing spend.

Relevance replaces frequency spam. Instead of three push messages a week, one that fits. Opt-outs go down, redemption goes up, and the brand feels like a good local: it knows its guests.

Platforms are reach, not relationship. What matters is the ratio: how many orders belong to you, how many to the platform? Steering that quota is a leadership task, because it determines margin and data access.

Your own channel stands and falls with operational integration. Orders have to reach till and kitchen without detours, otherwise the advantage collapses at peak times. The connection is the project, the frontend is the finish.

Every direct order builds your guest profile, every platform order builds theirs. Over a year, the channel split decides whether you know your guests or rent them.

Compare a year of platform commission with the cost of your own ordering channel. At a relevant delivery volume, your own route usually pays off faster than the next commission increase arrives.

Direct benefits need not be expensive: exclusive products, earlier promotions, saved waiting time. Everything the platform cannot reproduce makes your own channel attractive without sacrificing margin.

A ticket that is 15 to 30 per cent higher is not an optimisation. It works like a price increase without raising prices. Digital ordering improves guest experience and margin at once.

Kiosk, app and web have to draw on the same product and price logic. Otherwise the team maintains three menus. One leading system for menu and prices, every channel reads from it.

Digital orders deliver basket data at guest level for the first time. What gets combined, what gets left, what works as an upsell? That data makes assortment decisions measurable.

The kiosk case rests on two effects: a higher ticket and relieved staff at peak times. Both are measurable in pilot sites within weeks. Ask for the pilot, not the study.

Ordering digitally means ordering calmly. Images, extras, no queue behind you. With recognition and favourites the experience gets better, not more anonymous.

At the leading chains half of all visits run through loyalty. There the programme is the operating core of the guest relationship. Treat it as a stamp card and you leave that position to the competition.

A loyalty system is a rules engine with connections to till, app, kiosk and kitchen. The complexity sits in the interfaces, not in the points. Cut well, it runs alongside the existing estate.

The programme is your sensor. It links anonymous receipts to real guests. Only then do frequency, basket and churn data emerge at person level.

Adjusted for self-selection, 18 to 30 per cent more spend is the most robust figure in the industry. Measure the programme against that range, not against the cost of the free products.

Reward behaviour you want: the off-peak visit, the app order, the new product. That way the programme steers utilisation and assortment instead of handing out discounts.

Location and investment decisions are only as good as the data beneath them. A shared data model across sites and channels is therefore the basis of the expansion strategy, not an IT project.

Tills, apps and platform exports deliver different formats at different intervals. The task is a defined data flow into one central model. Built once, used daily.

Start with the question that really occupies management. What is an app guest worth, what a platform guest? For that one question the sources get connected. The foundation grows along real decisions.

Without a channel comparison, budget follows gut feeling. Usually into the platform growing loudest instead of the channel with the best margin. A shared data model ends that.

Campaigns across sites can only be judged when the data is comparable. Then it shows per location and channel which promotion brings frequency and which only hands out discounts.