The problem
Most customers see the same home page, the same films and the same offers, whatever they’ve booked before. That wastes the most valuable thing a booking platform has: what each customer actually does. Someone who always books IMAX on a Friday, with a large popcorn, shouldn’t have to search for it.
Personalisation goes wrong in two ways. It’s either a separate tool bolted on, which never quite reaches the checkout where it would matter, or it’s so slow that recommendations arrive after the customer has decided. It has to work inside the booking, in real time, in every market and language.
VOX Cinemas: recommendations in the booking
For VOX Cinemas, across eight countries in English and Arabic, we integrated machine learning and AI into the live booking platform:
- Film recommendations and personalisation (2019), presenting films to each customer based on their behaviour and history
- Food and drink recommendations (2021), with a personalised quick-add component inside checkout, so the suggestion is one tap from the basket
- Concessions ordered by local habits, cinema by cinema
- Targeted offers by customer segment, driven by the SHARE loyalty programme’s APIs
- Braze (2021) and Braze in-app messages (2024) for personalised messages across web and app, with Tealium, Firebase and Google Analytics underneath
See Food and drink ordering and Loyalty, memberships and subscriptions.
ITV: personalisation at scale
For ITV we built the User Data Store behind itv.com’s personalisation and ad targeting. It collected user behaviour alongside a site with more than 8 million unique visitors a month, asynchronously and fault-tolerantly, so it never slowed the site down.
- Behaviour, explicit preferences and profile data were grouped into segments.
- The CMS used the segments to return the most suitable content, and the ad server to return the most relevant advertising.
- It drove features such as recommendations based on past behaviour, “show me more or less of this”, and most viewed, rated and commented.
- Holding the raw data in-house let ITV build its own analytics, at a fraction of the cost of buying custom data feeds.
Everyman: the app knows where you are
For Everyman Cinemas, the app switches to an in-venue welcome when the customer is within 50 metres of the cinema, opening the in-venue features. After the film, a notification brings them back for extra content.
How it works
- Built into the booking. Recommendations appear where the decision is made: on the film list, in the seat map, and in checkout.
- Real time. Suggestions are worked out at the moment of the visit, from the customer’s history and what they’re doing now.
- Local. Recommendations and offers reflect the country, the cinema and the language, not just the customer.
- Joined up with loyalty and CRM. Loyalty data, Braze, Movio and analytics feed the same customer view, so offers and messages agree with each other.
- Consent first. Personalisation runs within the customer’s consent choices, with a consent platform such as OneTrust (integrated at VOX in 2023).
We use proven recommendation engines and customer engagement platforms where they fit, and build the integration, the rules and the customer journey around them. Read why in Beyond the AI hype and The digital sommelier.
Beyond cinema
The same approach suits any business with repeat customers and a lot to choose from: theatres and venues, theme parks, gyms, restaurants and retailers. The question is always the same: what is this customer most likely to want next, and can we show it before they go looking?
Related
- Services: Automation & AI, Software Development, Mobile App Development
- Solutions: Food and drink ordering, Loyalty, memberships and subscriptions, Gift cards, vouchers and partner offers
- Case studies: VOX Cinemas, ITV, Everyman Cinemas
- Insights: The digital sommelier, Beyond the popcorn
Frequently asked questions
Do we need our own data science team?
No. We integrate proven recommendation engines and engagement platforms, and connect them to your booking and loyalty data. You decide the rules and the offers.
Does personalisation slow the site down?
It shouldn’t. At ITV, the User Data Store ran alongside a site with more than 8 million visitors a month with no impact on its performance, and at VOX recommendations are worked out in real time inside the booking flow.
Can recommendations include food and drink, not just films?
Yes. At VOX, a personalised quick-add component suggests food and drink in checkout, ordered to reflect each cinema’s local habits.
