Who watches Unbreakable Kimmy Schmidt, the Netflix Original sitcom now in its third season? It’s a comedy set in New York about a 30something woman rebuilding her life after being held by a cult leader for 15 years in an underground bunker. The central themes revolve around the the resilience of women, the power of positive thinking, and not letting the past dictate your present and future – all set within the backdrop of the extremes and variety of New York life and society.
So why on earth have I recently binge-watched my way through 2 seasons? I watch sports, news and music documentaries, not US sitcoms about female resilience. Perhaps that’s why the Netflix algorithm never pushed Kimmy my way. Nothing about my personal data or historic viewing said you really need to push this show to Dan – he’s perfect fodder for this kind of stuff. And now I’m recommending it to who knows how many people through this post!
So if the Netflix algorithm didn’t track me down, how did I find out about Unbreakable Kimmy Schmidt and her hilarious antics? The answer is through a personal recommendation from a friend. My friend knew that I was doing a lot of work on people power and the way that positive thinking can shape positive outcomes, and she messaged me to say I should check Kimmy out. She even included a link to Netflix in the message to make it easy for me.
I then got to thinking about how Netflix recommends content to its subscribers – and it became obvious to me that they are missing a huge opportunity.
Netflix’s core proposition is simple watch what you want, when you want, on any device, and for a relatively unnoticeable monthly subscription. ┬áThey have made a virtue of convenience over content and are now increasingly shifting the balance of content from catalogue to originals. This means it is becoming increasingly important to have powerful recommendation engines.
Netflix makes a great play about its data driven recommendation engine which appears to make recommendations based on:
- historic viewing habits
- responses to previous recommendations and series imagery
- star ratings that customers apply to content they know or have seen
- demographic and personal information
- any information that can be gleaned from the social features that Netflix offers, which consist mainly of links to Facebook
The key point here is that the recommendation engine is based on customer data and algorithms to infer likely future preferences, and then recommend accordingly through each subscriber’s unique home screen.
Whilst there will undoubtedly be a strong correlation between fans of, for example, The Usual Suspects and House of Cards (spoiler alert: Kevin Spacey is the common link), this automated method of recommendation is missing the human element. As a result, the algorithmic approach has a number of blind spots, including:
- Individual tastes are highly complex. It is possible to be a passionate horror movie fan, or a casual romantic comedy viewer, and enjoy UK Channel 4 comedies. It is harder for a data driven algorithm to identify this complexity than it will be for your friends who will be very aware of your weird eclectic tastes.
- Social recommendations from friends are more trusted and more likely to be acted upon.
- Friends are more likely to know pieces of unique information about you that may encourage a recommendation – and one that you are likely to take seriously because of its personal and highly targeted nature. Just as I did.
Netflix does recognise the power of personal recommendations because it has a social feature that enables a subscriber to recommend a Netflix programme to friends on Facebook. But that’s it. There is no incentive to do so, no tracking of whether a friend responds to the recommendation, no alternative other than Facebook (in the UK service at least) and it requires that a subscriber agrees to Netflix linking to their Facebook profile, friends and preferences.
What Netflix, and indeed any other over-the-top (OTT) TV streaming service, could easily create is a personal recommendation engine to work in parallel to the algorithmic ones. Technology exists already (I’ve seen it) which could deliver the following:
- Tools to allow users to recommend content to friends over multiple social platforms and communication networks
- The ability to know if a friend responds to the recommendation (attribution)
- The ability to reward a user if their recommendations are acted upon. (Eg earn Netflix points for successful recommendations and trade points for money off subscriptions or access to programme-related goodies)
- A record of which programmes are recommended the most and, importantly, which are viewed more as a result of recommendations. Such data can inform future buying/investing decisions.
- A record of which platforms people share into and which platforms yield the most activity. This data can inform future marketing activity.
- Simple member get member features to incentivise existing subscribers to recruit new subscribers from their networks.
Such enhancements to a streaming service like Netflix would provide powerful tools enabling human interaction and people-power to drive the recommendations and uptake of service and content. In this way the recommendations will be more accurate, more trusted, more actionable and lead to a more enjoyable and social experience for users and their networks. This should deliver a cheaper and more effective marketing approach than existing methods. It might also reduce the amount of annoying push notifications and e-mails that algorithmic recommendation engines inevitably generate.
OTT services that harness such people-power will gain advantages over those that don’t. Not only will the marketing and awareness job be willingly and efficiently performed by an army of satisfied and networked viewers, but the increasing interaction with networked people will open up new and unforeseen ways to evolve OTT streaming as it moves further away from traditional broadcasting.
Just like Kimmy Schmidt, Netflix shouldn’t just rely on the past to dictate its future – it should also trust in the power of people to help build the future of TV from its solid foundations.
® Dan Allen