AI and Video Marketing: It's Not Just for the Major Brands
According to eMarketer, over 62 percent of internet users around the world will watch digital video content in 2017. Cisco believes that by 2020 video traffic will account for 82 percent of all internet traffic. With numbers like those, video marketers must face the reality of discoverability. How are they supposed to ensure their content gets seen by the right eyes at the right time amidst the ever-growing video landscape?
The answer could be in artificial intelligence (AI). While major brands and publishers are already using machine-learning technology to deliver optimized viewing experiences to audiences, small- to mid-sized video marketers can also capitalize on this solution to discoverability and user experience as the market grows.
The Importance of AI in Video Marketing
Iris.TV, a video automation and programming platform, knows the power of AI first-hand. The brand's customers generally see a 70 percent increase in video views, a 62 percent increase in views/session engagement, and an impressive 10 percent decrease in bounce rates. On top of that, Iris.TV's 10 biggest clients saw an average 400 percent increase in video views in 2016; one publisher saw a 465 percent view increase in just a week.
All of these numbers are because Iris.TV's clients used its AI platform to serve personalized, relevant videos and ads to audiences.
"AI/machine learning is going to be a huge equalizer for small- to mid-sized brands starting at the end of this year," says Robert Bardunias, co-founder and chief revenue officer at Iris.TV. "Brands are going to start to realize that AI and machine learning help augment the work normally done by a team of analysts."
"AI can provide a level of insight not available with general meta and share data, and removes a lot of the guesswork that goes into video targeting," explains Fritz Brumder, CEO and co-founder of live streaming platform Brandlive. "AI technology allows marketers to create multiple variations of videos by use cases, audience, message, product, and performance. AI helps marketers better predict outcomes, and makes it possible for brands and publishers to make recommendations based on interests and interactions."
So, if AI is that powerful, why aren't more video marketers using it? Brumder believes it's because the "scope of AI is intimidating," and that "we are in a walk before we run state." All marketers have access to the data they need to make AI work, he says, but it's conceptualizing and taking the next step that's daunting.
"We are all control freaks. Marketers even more so," Bardunias adds. "The truth is that as much as we may pretend to be trailblazers, no one wants to be the guinea pig. The problem is that AI has evolved; it's just that no one realizes it. It's cool to say that you are working with AI, but the moment you tell someone to divert some budget to it, they give you the side-stare."
However, plenty of larger media companies and agencies have already begun putting more of their marketing budgets towards AI implementations for 2017, Bardunias notes, and that "will only help speed up the learning curve that will help align AI metrics and marketers' KPIs."
"The real fun is coming in Q2/Q3 when more and more video marketers realize that a lot of the foundation has been laid," Bardunias says.
Tips for Video Marketers Applying AI
Video marketers of all size can benefit from an AI-powered strategy, even more so now that the barrier of entry is becoming less expensive, less complicated, and less abstract. As such, video marketers looking to implement AI elements should follow these tips to use it well:
Always define the goal: Don't put too much trust in AI's supposedly inherent ability to judge what viewers want to watch next, Bardunias warns. Without proper parameters, even AI has flaws. Marketers who implement the technology and let it run the show unguided risk missing or skewing their marketing goals.
"AI is an amazing tool to leverage some of the data sets that are in the market, but depending on the KPIs, not always a broad-based solution," Bardunias says. "The key tip would be to always define the goal for any new technology before you launch the campaign. The days of setting and forgetting or just blanketing and hoping that the AI decides the best course is not always the answer. A system is only as good as its inputs, and in video marketing that means people!"
Know the target audience. On a related note, video marketers must also have a solid grasp on who they're trying to reach. "AI is undoubtedly a powerful technology, but in order to receive appropriate outputs, you must have quality inputs," Brumder explains. "You can't just scrape the bottom of the data barrel and expect machine learning to magically turn it into gold. It's important to have a complete, thorough customer profile, with data from several different sources."
Brumder adds, "It's essential to paint the picture" for AI to make its best recommendations. This means video marketers must pull data from a wide variety of sources, not just their own websites or YouTube channels. Social media engagements, mobile usage, and similar benchmarks should also be primary inputs for any AI implementation.
Embrace the power of niche content. If video marketers are concerned about their brands being too small or niche to use AI, they should think again. In reality, AI can save such marketers hours upon hours of research and analysis time, a luxury only big brands with dedicated teams might have. This, in turn, helps video marketers guide their strategies, not micro-manage them, all while the views and engagement on their content skyrockets.
"With AI, in some cases, the smaller or more niche the brand, the greater the opportunity to be like marketer fruit flies!" Bardunias explains "If you know your target market/audience and where they live and play, AI streamlines the rollout process. That means more data to play with and turn into actionable intelligence."
Bree Brouwer's article first appeared on OnlineVideo.net