The 5 Key Ingredients For Better Entertainment Recommendations

by Daniel Rovira, CEO of itcher

You’re back home after work. Dinner is over, the kids are in bed. You have 2 hours to watch a movie or TV show before you head to bed, where you’ll drift into sleep and awake at a hazy 6am the following morning, only to do it all again. Those 2 hours of dedicated TV time are precious. And in a world with an almost infinite number of titles, where the last thing you want to do after a long day at work is spend time searching, becoming increasingly frustrated, it begs the question — how can we choose the best movie?

The Key is Discovery

There are many titles in the entertainment world and without any way of navigating through, it can be easy to get lost. There are currently 130,000 movies to choose from. That’s of 282,000 hours of viewing time, which translates into over 30 years worth of movies. Add to that almost half a million music albums and over 2 million books and it’s easy to see why ‘too much choice’ is becoming a real problem.

In this world of almost infinite titles, discovery is key.

The 5 Ingredients for Perfect Recommendations

It takes a lot to get perfect recommendations and with so many titles available, it’s important we get it right. Here are the 5 key ingredients detailing what a recommendations service really needs to provide accurate and meaningful recommendations to it’s users.

Independent

Rely on recommendations from a streaming service? It’s limited. As mentioned in my previous article ‘Why Netflix and Amazon Give You Miserable Recommendations’, streaming services will only recommend titles they can offer you.

Perfect recommendations come from an outside source, one that is unbiased and covers all platforms. That way, you know your recommendations are a product of your taste and not a product of a promotion.

Covers All Entertainment

The perfect recommendation engine should cover all categories of entertainment. From movies to books, TV shows to games and, of course, music.

The key to great recommendations is cross references to everything you’re interested in. That means your books should inform your movie choices, which can then inform your gaming choices and so on. Your taste is continuous, so your recommendations should be too. The more data your recommendation service uses, the better your recommendations will be.

Personalised

What good are recommendations that aren’t tailored to you? Taste varies enormously. Just because you like one genre of movie, doesn’t necessarily mean you’ll like another. This is where streaming services so often fail — there is a distinct lack of personalisation. Not all horror movie fanatics, for example, enjoy thriller movies. Recommendations, to be accurate, must be personalised to each individual’s taste.

Technology

Your recommendations should be available wherever you need them. On your mobile, on the go, on your tablet, on the big screen at home, even your Google Home device. Accessible, flexible recommendations that can be browsed at nearly any point in your day.

A Detailed Rating System

Having a more complex rating system means your recommendations can prepared in a more specific way.

Data input needs to evolve from binary to more gradual metrics. A ‘like’ or a ‘view’ is a good indication but does not say how much you liked it. Was it OK? Did you love it?

A more sophisticated rating system means your recommendations have been looked at closely. They haven’t just been recommended because you watched something with a similar tag, but because it’s made it through the filtering process and has been presented to you as something you’re likely to enjoy.

The Future of Entertainment Discovery, Now

these ingredients in mind, a new breed of media companies will emerge, and it will sit on top of the content owners and the content distributors. It’s purpose will be to help users to find and discover relevant content. And, as AI technologies become readily available, this is no longer the future of entertainment consumption, it’s the imminent present.