(This post originally appeared on Inc.)
One of my favorite restaurants is doing something that might upset some people. But I love it, and anyone in the restaurant or retail business should take note.
Outback Steakhouse–my go-to place for getting a great steak, fries and salad for around $30–announced that it will be monitoring customers with discreetly placed cameras at one of its franchises and then using machine learning (artificial intelligence or AI) to better understand their behavior.
The company is using a product called Presto Vision, a tool built specifically for restaurants, that, according to a statement, “uses discreet cameras placed in the restaurant lobby and other areas where hosts, staff, guests, and other individuals are automatically tagged in real-time and their motion analyzed.”
The technology will be piloted an Outback Steakhouse location near Portland, Oregon, where it will be focusing on “lobby analytics,” which means gathering data on hosts, waiters and customers so that wait times, cleanliness and customer satisfaction levels can be better gauged.
Anyone who’s had to endure a wait at an Outback on a Saturday night knows exactly what I’m talking about. The restaurant can get packed at times (which comes as no surprise to me because it’s delicious and affordable). But that popularity has its downsides and unfortunately wait times and general lobby chaos causes some prospective customers to leave dissatisfied. Outback, using this AI technology, is trying to address that and other challenges.
Ultimately, the technology would be used to video monitor all areas of a restaurant’s operations, including back-of-house, exteriors and internal dining areas.
How does this work? Presto’s algorithms are able to interpret events occurring on the captured videos–a messy table left too long, a disgruntled customer, a growing crowd waiting to be seated, a server’s activity–and turn those events into data which can be extracted and analyzed by management. As a result, performance metrics can be established and evaluated. Managers can be alerted to problems before they become bigger problems, coaching and training opportunities can be identified and fewer supervisors would be needed to monitor more areas. Using this information, it’s hoped that management can better forecast guest experiences and anticipate issues.
It’s great stuff, except for the obvious concern: security.
Some people aren’t crazy about being watched and analyzed. But both Presto and Evergreen Restaurant Group, the Outback franchisor testing the system, say that any video data captured is stored “temporarily for analysis” and then deleted after 30 days. Both companies say that that “no personally identifiable information is tracked or recorded” and that the system is not unlike similar systems being deployed by Amazon’s automated grocery chain Amazon Go as well as highway toll booths, traffic monitoring systems and imaging technologies that are growing in use by manufacturers and medical device firms.
If you’re in the restaurant or retail business, pay attention: This is your future.
Yes, there will always be a security concern. But if implemented correctly and carefully, the benefits from video and AI technology like Presto’s will significantly outweigh those concerns–and contribute heavily to future profitability.
“E-commerce websites have always had detailed analytics on how customers navigate their sites, but restaurants never have had access to this information in their physical stores,” says Rajat Suri, the company’s Founder and CEO. “With this product, restaurants can now have access to critical insights on how their stores actually work. This helps them provide better service, operate more efficiently, and reduce overhead.”
As an Outback Steakhouse customer who has easily eaten at more than 50 of its locations over the past 10 years (Umm…. I travel a lot and like to eat steak), I say: Bring it on. Go ahead and video me, track me, analyze me. You can’t make your steaks any better than they already are. But there’s always room to improve elsewhere.
