Most people don’t think twice when they open a mobile app and order food from a menu. They don’t have to think much about it because, in the case of Chick-fil-A Inc., the app is doing a lot of the thinking for them.

The fast-food restaurant chain, which specializes in highly popular chicken sandwiches, turned to the product analytics platform Amplitude Inc. approximately two years ago to help engage and retain customers. Using machine learning to develop predictive analytics, Amplitude was able to create an application engine for Chick-fil-A that customized a menu based on customer behavior and preferences.

“As new items get introduced to the menu by Chick-fil-A, you see the ones most relevant to you based on predictive affinity and all of the machine learning that we’re doing in the background,” said Justin Bauer (pictured), senior vice president of product at Amplitude. “The real beauty is they are able to configure all of this by a marketer through a simple user interface. This did not require an army of data scientists and engineers. They are able to use the Amplitude platform to build out this entire experience for their customers.”

Bauer spoke with John Walls, host of theCUBE, SiliconANGLE Media’s livestreaming studio, during the AWS Startup Showcase Event: Innovators in Cloud Data. They discussed why a personalized approach has been effective, how a focus on user behavior yields the best results, creating a self-serve approach to modeling without the need for major investment, and the role of emotion in building a connection with customers. (* Disclosure below.)

Major brands sign up

Amplitude was launched in 2014 as a real*time analytics platform. The company has grown its business to now power more than 40,000 digital products across over 180 countries. In addition to Chick-fil-A, Amplitude’s clients include Microsoft, PayPal Inc., Adidas AG and Capital One Financial Corp. Major brands have been drawn to Amplitude’s technology because the firm’s machine learning capabilities help shape highly personalized campaigns that can determine the likelihood a particular user will purchase or churn.

“It’s been proven that customers respond better to digital experiences that are more personalized, that are more relevant for them, and frankly just save them time,” Bauer said. “We see a big impact on a company’s bottom line if they’re able to deliver a more relevant customer experience. Personalization matters because it actually works.”

There are a number of different ways to create personalization models. One common approach is through demographics, using a series of questions directed at a particular user segment to determine baseline factors, such as gender or location.

“Most companies are actually doing demographic personalization,” Bauer said. “Are you a male or female, or do you live in a city or a suburb? The reality is that light segmentation is not really that effective. Do all women who live in a city behave the same? Obviously not.”

Amplitude’s approach is to focus on behavior as the best predictor for future action. That takes a great deal of data, architecture for tracking real-time user behavior, and machine learning models to determine which factors will result in the most optimal outcome for a client.

“We work with a content company that has millions of different articles, and they want to figure out the right article to put in front of you,” Bauer said. “That’s just not possible to analyze by hand nor actually orchestrate that in real time without leveraging machine learning. With new advances in supervised learning models, we can do that and generalize them for our customers.”

Understanding customer journey

Amplitude’s model also has appeal because not every company can afford an army of data scientists or engineers to build the machine learning algorithms necessary for success.

“What hasn’t happened is a self-serve way of doing this so that it doesn’t require massive investments in technical resources,” Bauer explained. “We’ve built a database specifically catered to helping you understand the journey of a customer across all of the different platforms and channels. You have to start there at a foundational level. That’s a big part of our secret sauce.”

The company’s democratization of data is designed to fill a gap it has identified in how businesses leverage the information gathered. Amplitude surveyed over 350 product leaders and found that only 20% of product teams had access to product intelligence tools with behavioral insight and 69% had to wait anywhere from a few days to a week to gain answers for simple questions about user behavior.

“You can’t personalize a journey if you don’t even know what your users are doing to begin with,” Bauer noted. “In and of itself, this is not a machine learning problem first. It’s a trustworthy data problem.”

Despite the advances in data collection and supervised learning, companies seeking to fully understand user behavior must be able to combine science with emotion. Personalization is also about knowing the emotional state of a consumer and being able to channel the best approach at exactly the right time.

“There is the art of understanding the emotional states that users are in,” Bauer said. “You’re still going to talk to customers; you’re going to understand them and what their different need states are. But this is then taking what you built as a great product and optimizing that. It is both about the art and the science coming together.”

Watch the complete video interview below, and be sure to check out more of SiliconANGLE’s and theCUBE’s coverage of the AWS Startup Showcase Event: Innovators in Cloud Data. (* Disclosure: Amplitude Inc. sponsored this segment of theCUBE. Neither Amplitude nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE