How Netflix Predicts What You Want to Binge Watch
Personalization is becoming a standard part of daily activities, from the songs on your morning playlist and the news articles you see, to the digital ads you are served and the product, book, and (online) friend recommendations you receive. Behind all of these experiences is a blend of man and machine, taking in customers’ information and continuing to refine and provide an improved and relevant experience based on those insights.
With an increasing amount of data available to marketers and the growing number of touch points individuals have with brands, tailored experiences are no longer a nice-to-have; individuals expect brands to understand what their needs are and deliver on them.
Below are a few examples of unique and inventive ways companies are blending man and machine to provide accurate and impactful customer experiences.
Pandora uses years of data + a human ear.
Pandora not only has a 10-year-old collection of millions of users’ listening patterns, but it also uses a sophisticated classification system called the Music Genome Project. A team of trained musicologists – who are first required to pass a selective screening process and complete intensive training in the Music Genome’s precise methodology – analyze songs using up to 450 musical characteristics. These qualities and attributes make up a song’s musical identity and are each assigned a value that ties into an algorithm. This algorithm enables Pandora to deliver a personalized station, based on an individual’s musical preferences, that serves music they are most likely to love.
Netflix blends film experts + consumer behavior.
Netflix also uses a form of machine learning to generate recommendations based on viewer behavior. Each year from 2006 to 2009, the company held a competition for the best collaborative filtering algorithm. Participating developers across the country created algorithms that predicted grades based on a training dataset and contributed to the algorithms that Netflix still uses today.
Similar to Pandora’s tactics, Netflix hires film experts to hand-tag movies and shows with specific attributes. From there, the company tracks what users played, rated, and searched for, as well as when they used Netflix and on what device. Combining product information with tracked consumer behavior gives Netflix the ability to recommend shows tailored to an individual’s viewing tastes.
Amazon provides relevant recommendations based on past purchases.
Of course, a discussion about a personalized experience wouldn’t be complete without recognizing the online retail giant, Amazon. On the simplest level, Amazon bases recommendations on past purchases, shopping cart items, rated and liked items, and what similar users have purchased. Amazon describes this process as “item-to-item collaborative filtering,” and the company has used this algorithm to heavily customize the user experience on its site. This algorithm works to show applicable textbooks to college students, baby products to new families, the latest tools to avid gardeners, and other relevant product information to everyone in between.
Transera matches customers to the most compatible agents.
Taking things offline, we’ve all talked to customer service agents on the phone, yet most people don’t know that data is sometimes used to ensure optimal sales success. Transera, a cloud-based call center solutions company, partners with businesses like The American Red Cross and Aon to ensure each company makes the most of every sales opportunity. This is done by matching individual customers with the agent to whom they are most likely to positively respond. In essence, Transera determines what kind of agent will be the best match for an individual customer to heighten the chances of completing a sale. According to Transera, they “use historical performance data to statistically determine which agents sell best under which circumstances and which customers are most likely to buy when paired with a certain type of agent.” So when you speak to a representative, it’s possible it’s not a random pairing; your past actions may have indicated you’d be best matched to that specific person.
Pandora, Netflix, Amazon, and Transera all have one thing in common: their sophisticated use of data, blended with a human touch, creates relevant and timely recommendations for each customer. During a time when developing a personalized experience for a customer is becoming less of an option and more of a requirement, using data effectively helps companies break through noise and stand out among the competition.
For more information on how you can incorporate personalization into your marketing plans, or enhance your current strategy, reach out to Rise.
You might be interested in:
The state of machine learning today is reminiscent of baseball scouting in 2002 when Billy [...]
The marketing landscape is fast-paced and constantly changing. As we begin 2016, a few Rise [...]