DesignSpy
AI
AI behind the success of Pinterest
Pinterest reported $1,7 billion in revenue in the year 2020 that makes 49% more than last year. Monthly active users worldwide reached 416 million and are growing right now as we speak. Pinterest's share price rose from $20 (in early 2020) to $85 as of February 2021. So what's the secret?
Pinterest is the place where people come to plan their future. And during the period of Covid - 19 most people got stuck at home, so people started to look for ideas and inspirations more actively than ever before to plan their futures, what will they do or buy after this all will be over. But the real success behind Pinterest stems from its ongoing curation of flawless personalized content and visual experience and of-course pinning.
Pinterest success is based on flawless personalized content and visual experience
Have you ever wondered how Pinterest is able to recommend, say, furniture or home decor in accordance with a particular taste of the user? This is all undertaken through deep learning, and its subset adopts neural networks for simulating the brain faster for data analytics and to instruct computer models. Understanding the goals of a specific search allows the platform's deep learning models to serve personalised results.

Speaking of the neural network, you can take a look at our blog on What are Neural Networks and how do they differ from Machine Learning and Deep Learning?

How does Pinterest use Deep learning and Neural network?

No humankind can go through hundreds of millions of images every day, so Pinterest implemented AI technologies to help us find exactly what we are looking for. Here are 4 categories of how it is using the AI.
  • Visual categorization and management
    Particular metadata is being analyzed to complete this task taking into consideration for example pinboard name or date from the website where the picture is taken
  • Image Resemblance Identification
    Neural Network and deep learning is being leveraged for processing loads of image searches each day, helping users in finding content that resembles the images they've pinned before
  • Concentrating on individual taste and actions
    Deep personalization is the key goal for Pinterest not the general amount of likes and follows that other social media platforms focus on
  • Heading deeper than pictures
    It analyses captions of the pin and also focuses on items being pinned to the same virtual board.

In addition machine learning is also being used for detecting and hiding the content that demonstrates and promotes self-harm. It also abolishes harmful content at a 3 times faster pace than before.

Thanks to the highly personalized content Pinterest has grown at a fast rate in recent years but still has plenty of opportunities to capture more revenue and improve upon its user experience.
Made on
Tilda