Coffee, coding, cinema and making data-informed decisions Part 1/2

“That’s one small step for man, one giant leap for mankind”- Neil Armstrong

It is a calm afternoon in the heart of Indian capital. I have taken some time off of not my so busy schedule to pen down an official release note of our next set of features, which are few, yet a big step towards our reason of existence: personalize entertainment.

For a small introduction, MyMovieRack is a platform to organize and discover movies &TV Shows. Simple, we help our users decide on what to watch in this world full mediums to waste, invest and kill time.

From pretending to being a ‘customer first’ startup

Coming back to our recent updates, it has been an exciting journey for the past six months. Though I agree, in the world of start-ups, ad-hoc ideation, making ‘business decisions’ on WhatsApp groups, no-documentation of products is excused in favor of agility and velocity, one cannot excuse the two principals:

• What do your customers/users really want?
  which can be put as What’s in it for them?
• What drives your motivation to build and manage the product?
  which can be put as What’s in it for us?

I wonder how would the world function if that farmer didn’t discover his funny acting goats?

To answer the above two questions, we went back to our $3 drawing board with our old partially dried out marker and a cup of cheap coffee. After an extensive primary research, analysis of product usage analytics and brainstorming on insights(deliberately avoiding any or all business frameworks I have studied in my MBA school), we came up with the below set of tasks:

• Letting our users discover movies and TV shows using how they #feel about them.
• Support for Regional/multi-language content/assets on the platform.
• Improving credibility of the MMR score, the score which defines the perception of a film, TV show relative to other films/TV shows.

Summing it up, we needed to make MMR a better platform for all of its users.

Getting into shape and getting the right gear

For a small introduction, we( a team of 3) are no more than commoners who are still trying to figure out their level in the Maslow hierarchy. Out of all the regrets and mistakes we cherish, we are proud of one, that is taking up Engineering( though everyone knows that in my country that’s the most obvious thing to do, one never has a choice in the first place). Our own hunger for movies and TV shows is complemented with our education background and we could easily define the skillsets we needed to gain to solve the above three problems. Solving problems using hunch and cold coding would definitely not work. It was time we packed our arsenal with techniques such as User Experience Design thinking, classification techniques under Machine learning and yes the principal of customer centricity.

Feature 17.4.1: Discovery Engine to explore movies by feelings

The first update in this release has been a unique & dedicated discovery engine for searching movies and TV Shows using feelings such as #mindBending #bestOfComedy #erotic which capture keywords more than conventional and stoic inputs such as genres or just the name of the film. We will invite you to try out the same here. Its a powerful tool. Be creative and use it responsibly. Hey! look what I found Dark Psychological Thriller Drama Movies. And here is another one Lighthearted Feelgood Chickflick

The search results are put in three brackets with further filter options:

• Recommended: These results are curated as per your likings, ratings, recommendations. Your personality has higher weightage over other factors such as popularity, average ratings etc.
• Latest: These are the results where the recency aspect of a film or TV show has higher weightage over other factors such as popularity, average ratings etc.
• Popular: These are the results where the popularity(number of ratings) aspect of a film or TV show has higher weightage over other factors such as popularity, average ratings etc.

So Whats the deal with this Discovery Engine. How is it different?

You may be wondering where do these feelings for each movie come from (OK now you may be)? Here comes our USP. We are a bit different than traditional movie rating system (actually a lot more). Instead of relying on 5-star rating MyMovieRack allows a user to rate the movie in terms of how one “feels” about it. To do it optimally, 8 most probable feelings are shown to users while rating. A huge amount of brainwork(sampling and clustering) went behind showing these 8 feelings that keep the right balance of providing users the choice of exploring the new options and exploiting the old ones which were used frequently.

Upto 8 optimal suggestions shown during the rating at MyMovieRack

So this Discovery engine is blend of our passionate MyMovieRack community, our machine learning algorithms and big data capabilities which makes it largest, ever-growing feeling based library for movies & TV shows.

Feature 17.4.2: Filter by User, Cast, Director and Language

MMR offers granular filtering to enable you to pinpoint movies, TV shows which you might be interested in to savor your niche taste buds.

And there is much more, all simple one click access…can anyone let you do it so easily?

So what are you waiting for? Get set Discovering!!! at here

Soon we will bring the next part of the same article dealing with spam ratings and how we tackle them. Stay tuned!

Update : Read part 2 of the same series.