Scaling the content quality to keep your users pinned

Dreaming about the profound habit disruptions generated by the biggest consumer mobile applications of our generation is what makes me as an entrepreneur always seek for a better understanding of our world. Studying our most primitive habits allows us to identify a pattern that could potentially constitute a powerful opportunity.

This understanding allows us to build for our market the product of its dream to answer one of its most primitive and authentic needs – ego for Tinder, need for non judgmental interactions for Snapchat, …

 

When it comes to building a product used by an entire world of passionate users, we can easily identify two major steps which are (i) finding the perfect product-market fit and (ii) being able to scale that product for millions of users around the world. These steps define the biggest challenges of consumer mobile applications in their journey to becoming a commodity in our world.

Growing from 10,000 to 1,000,000,000 users implies the development of strong and sustainable infrastructures to insure a consistent content quality that even has the ability to improve as the user base and interactions increase at an extremely fast pace.

I was beyond seduced by the presentation of Jure Leskovec about the use of machine learning at Pinterest, where he operates as Chief Data Scientist. My goal is to aggregate in this article a simple vision of the challenges that emerging technologies represent for the techniques that Facebook, Snapchat, Instagram, Quora or Tinder use to allow you to always discover the best article, pin, match or friend that fits perfectly to what you are looking for and to your interests.

In the first steps of their user experience, the performance of the model of the consumer mobile applications strongly relied on the information inputted by the user directly to get a good understanding of his profile and interests (age, gender etc…) and display him the best content. This would result with never ending forms and profile editions, repetitive modals that would popup as we navigate, asks for tags inputs, empty wall at on boarding etc…

With the need for extremely simple products and Facebook being the pioneer in the collection of data without complexifying the user experience, we entered progressively an era where the collection of data on our users must not constitute an effort for them but must occur naturally in the user experience with the collection of metadata and behaviors rather than raw standardized data. An important characteristic of this model of data collection is that the quality of the content displayed to the user throughout time improves to provide him with an increasing product value, and that the qualification of the profiles is increasingly accurate. This allows to remove any friction and to pin the user with the best content that he is looking for at any time of his lifetime value for the company.

Pinterest for example recently released a new feature to allow users to research similar images, Facebook with the open graph opens an unbelievable web of connections to the world, Quora succeeded in building the largest existing community of experts by tackling the paradoxal growth potential of expertise and number…

It is the emergence of new technologies such as image recognition, text data mining, analytics tools and many more throughout the time that constitutes an enabler for consumer mobile applications to improve this content quality process.