Beyond The POS: Going Further with AI Retail And Hospitality Data

June 4, 2019

Digital POS systems have been the heart of brick and mortar retail since replacing analog cash registers in the late 1980s. The automated reporting and level of data they were able to gather provided significant upgrades and gave operators a more sophisticated view of their shoppers behavior. Today, as merchants are facing an increasing volume of cash and credit cards that pass through every day, digital POS has become a necessity to retailers, restaurants and hotels, due to its powerful features.

Of course, for modern day POS systems, the software is no longer just about processing sales but comes with many other features including order management, employee scheduling, accounting management, customer engagement, etc. Even more crucial is the ability to offer a seamless omnichannel experience. With these types of implementations, a business can upgrade to a more efficient and streamlined service, all while allocating staff more strategically.

This information is incredibly valuable—it can help a hotel managers understand each individual that has entered the property. Based on the information, the manager can direct advertisements and improve the customer relationship.

The challenge though, is the activities and data that exist outside of the purview of monetary transactions. That could be crucial information for a hotel manager to maintain a high-quality service. Let’s take a simple example: clearing dirty room service trays from hallways. After a guest has received the food from room service, the guest will usually leave the tray in the hallway after finishing the food. How long does staff need to wait before cleaning up the leftovers in the hallway? How can we optimize the employee schedule to keep hallways always clear? We can imagine having a person on each floor waiting patiently for trays to appear, but that is an expensive solution to a simple problem.

The bottom line is, operators want to ensure that the leftovers are cleared in time but don’t want to over staff and create unnecessary labor costs.

There are dozens of these types of problems across a large hotel. And each need sufficient data to make sure that the staff schedule is reasonable according to weekly, holiday and seasonal effects. Take another example like the Hotel pool. Guests don’t pay for usage, so tracking via POS data is unhelpful. To begin to solve the problem operators must understand exactly how many people are at the pool during each hour of the day, and in what ways. Imagine the value in knowing precisely, on a minute by minute basis:

A good sample of such data points (for at least a month) would allow a good estimate of labor demand for each hour during the day, and how the demand varies during the weekday versus the weekend. Given a larger data sample for at least 12 months, we can be more confident in analyzing seasonal effects and holiday effects. Similar analysis can be applied to any blocks on the property that provide an integrated service rather than one-stop shop, such as valet parking, brunch buffet, bars, etc.

Previously, the only way to collect this level of behavior based data was meticulous human collection. (Literally sending a staff member to count the number of people in specific areas and their behavior.) This is costly, error-prone, and gives barely enough meaningful data to be useful. That is thankfully changing. With advancements in AI and computer vision, hotels can now use their existing security cameras and AI software to answer those questions, all while protecting guest privacy.

POS data is certainly a key to optimizing a hotel or retail store’s operations, but it doesn’t contain all the necessary information for to keep guests happy and staffing efficient. None of these can be done without accurate data acquisition and robust analysis. The AI and computer vision will bring about a sea change in how operators envision the store or hotel experiences of the future.