These New Facebook Features Are Increasing the Importance of Computer Vision For Marketers
On the heels of Zuckerberg’s Capitol Hill testimony, Facebook has revealed several key changes to its platform, which rely heavily on artificial intelligence and computer vision, in particular. These updated features aim to make the platform more privacy-aware for consumers and safer for advertisers. Here’s what marketers need to know:
1. ) Clear (Data) History
The Clear History feature will (allegedly) allow users to see the websites and apps that send Facebook data, delete that data from your account and turn off Facebook’s ability to store that information associated with your account moving forward.
While that’s all well and good, as a marketer and Facebook advertiser, it does set off alarm bells as to how this may limit the ability to deliver behavioral based ads to users and make retargeting difficult. This creates potential reach and performance issues, which ultimately can translate into lost revenue. If marketers can’t reach their audience will they go elsewhere? And how many consumers will actually use this new feature in the first place?
2.) AI Sorting In New Explore
Altering content algorithms, whether on the News Feed or in Instagram, has always been a touchy subject for the company given extreme reactions users have anytime something changes. So it was surprising Facebook announced the redesign of Instagram’s Explore page, particularly that the new Explore will sort content into relevant topic channels using AI.
The intention is to make it easier for users to browse and find the content they’re most interested in. As a consumer, I welcome this change as I currently find most of the content in the Explore tab not necessarily aligned with my interests. This being said, it remains to be seen how this will improve my feed as the actual results of the curation are dependent on the ever-learning AI behind the scenes.
What’s clear is that Instagram is betting on computer vision to do the heavy lifting, which is no surprise given the sheer volume of images and videos posted on Instagram every day. But even that is no easy feat. Computer vision is one of the most powerful and compelling types of AI that currently is leveraged across drones and connected vehicles among other things on the consumer world and will be incorporated more and more as a result of the increasingly number of images shared online everyday (currently estimated as roughly 3 billion).
3.) Instagram Bully Filter
Brand safety and toxic online environments is top of mind for brands and marketers as GumGum’s recent brand safety study discovered. Social platforms have been taken to task the past year with a steady stream of negative news around online bullying—of children and adults. It’s a top concern with publisher and brand partners who want to be where people feel safe, and if social platforms don’t offer that, then they’ll ultimately take their money elsewhere—but they haven’t yet. Brands are still pouring money into Facebook.
To its credit, Facebook is starting to tackle the problem with the Instagram Bully Filter. From a UX and the back end perspective, Facebook needs to train large data-sets that drive powerful image recognition capabilities. This is hard stuff and expensive. To power the Bully Filter takes a lot of AI tech—especially the use of computer vision.
This tech fills a gap that text-only analysis creates. Context is everything these days and content can be used as a proxy for audience, ironically the way it is used in broadcast TV and print media. When it comes to protection and brand safety, brands and marketers need to be hyper-vigilant of both images and text, particularly on a platform as visual as Instagram. This is a critical move for Facebook as it tries to repair its image with brands and consumers alike.
So, really Facebook is an AI company—or rapidly becoming one. This is line with the growing proliferation of AI investments from eco-system players, from Facebook to Snapchat to Pinterest. The investments these companies are making now will bring a whole new feature set to marketers in the future.
The Renaissance of Contextual Advertising
As access to consumers’ personal data becomes more scarce, marketers find themselves returning to a tried-and-true technique that a few short years...
The Problems With Pre-Roll
Throughout the relatively short lifespan of digital video, advertisers have tried myriad ways to make pre-roll more palatable. Yet it remains far f...
When Brand Safety Goes Bad
Here we examine a host of maladies that traditional brand safety methods can lead to—and how to find a balance between safety and common sense that...