Video was supposed to be the savior of the digital publishing industry.
It still can be, just not in the way we all expected. Instead of piles of capital, pricey talent, and expensive production operations this time around, breakthroughs in AI and editorial automation will bring about a video renaissance.
Think back to the mid-2010s (not so long ago – yet so long ago) – venture funding was pouring into digital media and tech giants were promising riches. Then somewhere along the way, the phrase “pivot to video” became an industry joke; trying to make money in the walled gardens became fruitless for many, and the ad budgets never came fast enough to fund the expensive studios and high-priced talent. The video boom seemed over before it started.
And yet, consumers keep watching more video – on pretty much every device. eMarketer forecasts that people will stream 33 more minutes per day on average this year compared to 2019. Aside from the red hot CTV sector, online video now accounts for 133 minutes of viewing per day among US adults, found the researcher.
Given the declines we’re seeing in linear television, it stands to reason that advertisers will pay a premium to reach a streaming audience. Yet based on our work with publishers, many sites on the open web do not produce videos or have a sizable video audience.
Following the ill-fated video pivot, publishers may be gun shy when it comes to investing in video. Or perhaps they are just ceding this lane to the social giants.
However, we believe that the digital publishing world has a chance at a second video boom – one that can be far more lucrative and sustainable – if publishers are willing to embrace a certain level of automation.
I can already feel it. Somewhere, an editor is clutching his or her chest, holding back screams. There’s no need to panic.
No one is turning the newsroom over to robots, putting reporters on the unemployment line. We see video automation as a boon to journalism, freeing up journalists to do more of their best work while bolstering the company’s profits thanks to hefty video CPMs.
Think about a digital news organizations’ current video output, for example. The lucky ones are attached to TV networks, so they have the built-in advantage of a constant funnel of fresh clips from which to choose – most do not.
This TV network-connected site likely produces some video for its social channels, which may require a team of editors to spend time reformatting, cutting down and annotating as many clips as they can for every feed. The original footage might be great, but the execution is laborious and inefficient at best. Could machines do that better?
What about “templated” video types, such as standard reports on financial earnings, or sports scores? An informational video is beneficial to consumers – but feels like an inefficient use of resources by highly trained producers. The time is coming very soon when software and AI can produce this kind of work.
The dream scenario is that media companies can enable their top journalists to work on great stories while producing much more video. Over time, readers get accustomed to watching short video clips in a faster, more digestible form. At this point, video consumption will become like the Flywheel Effect from Jim Collins’ book “Good to Great,” whereby viewership continually builds without having to exert as much effort.
On the business side, publishers will have more inventory to work with. They can come to advertisers with a legitimate video audience, one that is more inclined to see and become captivated by strong video ad creative. The publication makes more money, allowing management to hire more journalists, and consumers get better and more content too. Everyone should win.
It will take media companies, especially those that cherish the “purity” of journalism, some courage to overcome the stigma attached to the idea of automation. Yet, it’s worth looking to other industries for inspiration – and comfort. For instance, AI is being used to speed up medical research and even vaccine deployment. Closer to our industry, Netflix and YouTube’s content recommendation engines help humans discover content they love, without having to sift through endless choices.
Plus, the automated reporting wave has already started, and media firms like Bloomberg and the New York Times are experimenting in using technology to produce content. These are not companies that take journalistic integrity lightly – but they absolutely want to operate smarter and more efficiently.
This AI-based video revolution is not that far away, and we think it can fuel a second web video boom. The only pivot will be toward a much healthier industry.
Comments are closed.