The Future of Digital Guest Data for Restaurants

The Restaurant Data Challenge

Cloud-based POS systems enabled restaurants to access their data virtually anywhere, at any time. With the transition to more digital engagement and off-premise ordering expedited by COVID-19, restaurants now have the ability to collect data about their guests beyond the restaurant and from all of the different ways, guests engage with them.  However, collecting data and taking action before it becomes stale are two very different things. 

While the largest restaurant brands with multi-million dollar IT budgets are able to afford the labor and technology to turn data into actionable insights, most restaurants do not have that luxury. For small to mid-size brands, data is often spread across many different platforms - one for online ordering, one for loyalty, one for email marketing, and so on - and often in different formats. By the time that data can be collected and correlated, it is often too late and is more helpful for seeing what happened in the past than it is for driving real-time action that influences results. 

As on-premise dining continues to return post-COVID, restaurants recognize that improving digital engagement and commerce will continue to be critical to their success. What many restaurants may not realize is that ensuring their digital tech investments help them achieve their goals relies on their ability to leverage all of the data they’re collecting about their guests and make it actionable. 

(source: Hospitality Technology 2021 study)

The Status Quo

We’re seeing the industry move increasingly towards all-in-one digital guest experience platforms. Why all in one? The obvious benefit is a reduction in expense and the simplification of your restaurant tech stack. Over time, however, the real advantage is that the data won’t be spread across silos. Your CRM should contain your commerce data, guest demographic data, and even your marketing results; the same platform should provide the tools you need to make the data actionable. 

An all-in-one platform provides you easy access to the data, with the outreach tools built right in. Think about creating new customer segments and being able to target them in real-time with the right marketing communications.

The Future - Autonomous Digital Guest Experience

Cloud-based POS was the first (enabling) wave, and the rise in digital engagement and commerce was the second. The third wave will combine the two with machine learning to streamline operations, enhance the guest experience, and improve the bottom line.

Digital guest experience platforms should be learning over time using your own data. These platforms will look at things like what items are guests buying together, which items cause guests to increase visit frequency, what is the next logical product, and how a well timed offer can influence the buying journey.

One early application of machine learning for how restaurants is solving can upsell products based on their own data to drive incremental digital revenue. The best upsell models use a multi-level machine learning architecture and creates metadata before going down an additional level to create a recommendation model.

Restaurants already leveraging an all-in-one platform with built in upsell recommendations have seen some interesting results without any additional budget or operator investment. For one restaurant, the machine learning algorithms identified a segment of customers that bought two desserts with a strong correlation between a cookie and a marshmallow bar! Wing it On, another restaurant group leveraging the same platform, immediately saw a 4% increase in digital revenue driven by autonomous upsells. 

Over time, leveraging machine learning will only become more common and help restaurants increase upsell revenue without the attendant problems related to manual rules upkeep and replacement recommendations. While upsells are an interesting use case, the industry is just starting to scratch the surface of what is possible with a complete data set and sophisticated machine learning models.