1: Instagram is a photo-sharing app that has surpassed 500 million daily active users. In contrast, rival Snapchat has 229 million. Instagram allows its users to manipulate their photos with different filters and share them with friends and the greater Instagram community with hashtags and location tags. One thing that makes the platform so engaging for its active users is the personalized experience Instagram creates using artificial intelligence and big data.
With more than 70 million new photos posted to the app each day, Instagram has endless data logged from the ways users engage with posts. Instagram uses machine learning to populate its Explore tab, an area in the app where users can explore content on the platform.
According to Dan Toffey, Head of Community Lab, Instagram dives into vast amounts of data to understand how people use the platform and to create a curated experience for users. For example, Instagram is able to adjust a user’s feed based on likes, comments, follows, search preferences, and more. The app continuously learns from user behavior and adapts over time.
“We have this rich diversity of data that we’re able to use to look for constellations of accounts. Then we can go in with humans and look at it and understand what they’re doing,” Toffey said.
Instagram’s Community Lab team works to unearth insights from data trends. Big data is valuable but it is useless when a company doesn’t know how to use it. Mining big data for customer insights can take a lot of time and resources. Instagram prioritizes this and is able to drive product development and extract community insights as a result. Instagram strikes a balance between artificial intelligence and human curation to deliver personalized content as well as subject matter that may be outside of a user’s interests to create new opportunities for engagement.
Instagram also uses this data to sell advertising to companies that want to reach various user groups. Facebook purchased Instagram for $1 billion in 2012 with the intention to break into the mobile industry. With more people using mobile apps or accessing the internet through their smartphones, the use of photo sharing through mobile devices is increasing. This also adds to the data pool Instagram has to work with to offer highly targeted advertising.
Instagram also uses artificial intelligence to filter out spam posts and comments. The system, called DeepText, was created by Facebook to remove fake messages across nine different languages: English, Chinese, Russian, Arabic, and more. It has the ability to understand not just bad words but the context in which they are used. DeepText was trained by humans who sorted through messages and classified them as spam or not spam. It has expanded to also remove messages that fall into the categories of bullying, racism, or sexual harassment. While it’s not a perfect system, the filter outperforms static algorithms because it is always learning.
“The whole idea of machine learning is that it’s far better about understanding those nuances than any algorithm has in the past, or than any single human being could,” said Kevin Systrom, Instagram’s former CEO and co-founder.
“And I think what we have to do is figure out how to get into those gray areas and judge the performance of this algorithm over time to see if it actually improves things.
Because, by the way, if it causes trouble and it doesn’t work, we’ll scrap it and start over with something new.”
Above all, Instagram’s use of big data is aimed at enhancing the user experience to ensure users keep coming back for more. The influx of content and users on a daily basis require the platform to be nimble when it comes to learning from its findings and adapting the app.

Questions for Discussion
1. Describe Instagram’s process for populating content on the Explore page.
2. Why is it important for Instagram to create a tailored app experience for each user?
3. In what other ways could Instagram use big data to improve its platform?