The big topic in the world of TV Everywhere is content discovery. I can relate because I am a movie geek. Avatar to Zelig. Bogart to Willis. Altman to Zeffirelli. I’m a fan of them all. I was one of those kids who skipped school to stand in line on the opening day of The Empire Strikes Back. (I did it again for Pink Floyd’s The Wall - my algebra teacher saw me leaving the theater.) I even went to film school. I have spent more time with movies than any other type of entertainment. And now, that’s part of my problem. It’s getting hard to find good movies that I have not already seen.
While most consumers may not have watched as many movies as I have, they still want to find good new content on a regular basis. This is why content discovery is such an important challenge for our industry. The better job we do helping connect consumers to the content they want, the more we all win.
The IVA team has spent a lot of time this past year thinking about content discovery as part of our overhaul of VideoDetective.com. Video Detective is the perfect lab for us. We can try ideas and products there before they are offered to our clients. So let me use some of our discussions as a jumping off point for talking about the art of content discovery.
Crowd-sourcing Crowd-sourcing has been a hot concept for a couple of years, and there’s no question that it’s invaluable. Netflix and Amazon have pioneered the art of making recommendations based on the extensive database of activity of their users. But there are a lot of steps involved in getting it right. Crowd sourcing works for identifying the obvious choices, but it would be great to see personalized recommendations that will also find the forgotten movie gems I am always seeking.
A Question of Taste Knowing a consumer’s tastes in movies is critical to making great personalized recommendations. But how do you find out about taste without bogging consumers down with an hour-long questionnaire? Ask just a handful of key questions. What are their favorite genres, actors/actresses, directors, films of all time?
It is a safe assumption that if I like a particular director or actor, I will like more of their movies. Knowing a couple of movies that I think are my favorites of all time can guide me to other similar titles based on genres, directors, cast, etc. If your company can leverage Facebook, you can ask for consumers to provide you access to their profile information and their friend’s information. This could kick-start your content discovery engine with more titles that your consumers ‘like.
Recommendation Filters Of course, it’s not enough to just gather the data. What is needed is a relational database that captures this information and matches the data to other titles. With thousands of movies available today, there can still be a lot of content to wade through. So, you will still need to reduce the “noise” in the results to allow the best recommendations to float to the top. We’re working with a company called The Filter to help us improve Video Detective recommendations. Key ways to sort recommendations can be based on:
- Social Media – If consumers allow you to access their Facebook friend’s info, you can use that data to see what movies are most popular among that user’s network.
- Critic Ratings – Rotten Tomatoes and MRQE aggregate critical ratings and reviews and are a good vehicle for identifying better titles.
- Social Ratings – Flixster has a huge database of users rating movies.
- Awards – Movies that win awards such as an Oscar, Golden Globe, etc. lends credibility to the film’s credentials – and yours! Nominations and winners are easily found online in many places – including from IVA.
- Box Office Gross – Popularity is one of the best barometers of potential interest, and box office gross is a great indicator of popularity. BoxOffice.com is the industry’s premiere source for weekly grosses.
- IVA Charts – Our weekly charts of Most Anticipated Movies based on trailer places is a great indicator of potential interest for films that haven’t been released.
Changing Tides Consumer habits can change. Documentaries have gotten more popular since Michael Moore and YouTube. SciFi titles are bigger than ever, and girls play videogames. So even after you’ve captured your first round of data, you have to update it with “real-time” data. At VideoDetective we’re using what titles a user chooses to track in their profile just as Netflix tracks what movies you put in your queue or Amazon tracks your book purchases.
Closing the Sale Once the consumer gets to a title they are interested in, there are two things that are important to helping them make a final decision: data and trailers.
Data about the title will help inform the consumer about the movie. IVA provides the basics of director, cast, genre, rating, description, etc. If it’s a foreign title, they might want to know if it’s available dubbed or with sub-titles. We also work with a variety of top notch data providers including Rovi, Gracenote, Baseline, and IMDB. So depending on your target audience you can choose the level of detail you want to provide.
Now, I might be biased here, but trailers have been proven time and again to be one of the most important drivers for consumers to push the play/buy button. Trailers have been part of movie marketing since the beginning and now in the digital age play even a greater role. The “try before you buy” proposition is one of the oldest in marketing but is also one of the most trusted. Today, trailers build buzz, go viral, build anticipation and are clearly the best tool in movie marketing. Having them at the point-of-purchase is not only logical, it is critical to driving greater consumption.
Availability Once you connect someone to a title that interests him or her it’s helpful to let them know where to watch it. Movies are available across a wide variety of distribution channels these days, so release windows are important depending on how you structure your service. Interest generally starts with movies that are coming to theatrical, so for many of our clients the job is done once they’ve got theatrical recommendations handled. But today much viewing happens at VOD and streaming. If you are a general consumer site and not a retailer, letting users know where a movie is available helps to target results. Lastly, having Oscar nominated movies or top box office data lends credibility to your recommendations.
User Interface The last piece of the content discovery engine is the user interface (UI). Perfecting a UI is truly an art and there is much room for variation. Great UI is easy to use and visually appealing. It includes intuitive navigation, speed, and a multitude of other ingredients about which much has been written elsewhere. For movies, good UI includes using images such as posters, movie stills, or box art is a common, but powerful visual way to assist navigation.
With touch screens, voice-activation, motion detection, etc., the tool kit for creating the ultimate UI is changing rapidly. This is where the battle lines are being drawn. At any moment, a new app could emerge that consumers will latch on to, at least until another one comes along. Having a strong content discovery engine supporting the UI will best engage consumers.
IVA clients are choosing from the different items in our tool kit, and positioning the same items with different emphasis depending on who their target audience is. They are also bringing their own creativity and focus into play. This is making content discovery more robust in the marketplace because movie fans find movies in different ways.
Will we ever have a perfect recommendation engine? We will have lots of different ones that depend on your taste, and your tools - and that’s even better.