Media Publishing in the Age of AI – Back to Basics

Artificial intelligence is fast becoming an essential ingredient for an array of solutions across all industries – which promises to transform every aspect of our lives. While intelligence similar to (or perhaps even surpassing) that of humans may emerge in the not-too-distant future, it’s clear from our collective everyday experience we are not there yet. We are, however, firmly on the path. And there are steps we can take to create the brighter future we desire.

Daniel Harrison, chief revenue officer, IRIS.TV

Daniel Harrison, chief revenue officer, IRIS.TV

This is especially important when it comes to the media and publishing industry.

Trust is increasingly hard-earned

Concepts of truth and accuracy in reporting and storytelling are being tested as never before. AI driven newsfeeds have given rise to Fake News. And with Deep Fakes, content creation and manipulation have become less expensive and virality impartial to truth. Yet we the people feed the demon, as speed to market and sensation is rewarded, whereas quality, accuracy, and truth borne of patience, less so. Often these are a result of quick-fix solutions to core business model challenges, whereas figuring out the right model is much harder though enduring.

So how do we maintain and build a publishing business in this complicated high-risk high reward climate? How do we to capture the full benefits of today’s AI without losing control of what makes our brand (and our advertisers’ brands) unique, powerful and human?

Home is where you create value

For the past 20 years I’ve advocated for the adoption of advanced technologies, yet always calling attention to the need to first set the stage for the technology to most effectively achieve its goal while limiting the potential for adverse outcomes. Otherwise, the risks for adopting, then soon after sunsetting, the technology increases. With that, a burn of unfulfilled expectations (and empty pockets) lingers long across our teams and companies.

Without a doubt, the greatest step a media company can take in preparation for AI is to position itself to have a direct relationship with the user. One’s owned and operated properties (be they on desktop, mobile, OTT, or elsewhere) are the ground zero locations for capturing and developing these relationships.

If it’s an audience you can’t equitably monetize, can’t learn from, can’t leverage for improving what shows you cast, produce, distribute, and build your business from, is it an audience? Ask Defy Media how that worked out. Or Disney’s recent write-down of its stake in Vice.

It’s true that we can’t expect the O&O “we built it, so they will come” mindset to succeed on its own. We DO need to go where the audience is. Deals with YouTube, Instagram, Snapchat, are necessary and smart, especially if we are to start thinking about content and shows as “Omni-brands” that can and need to live everywhere and anywhere. But we must not primarily invest where we give the goods away. I walk into way too many meetings where this is the team that’s funded, not the guardians and builders of the house.

It is within the four walls of your O&O house that you are able to create a clearly defined and controlled experience for your customers around who you are, why you matter, the benefits of sticking around and returning. It’s your scientific laboratory – where else are you going to get such clear unfiltered signal to test what’s working and not, to inform content creation, distribution, licensing, business models, new products, and all other experience and brand defining investments?

Laying the foundation for AI

Getting the house in order is no easy affair, but its table stakes. This includes having a common framework for testing, deploying and learning. Centralization and standardization of assets and a common data model with dynamic taxonomies are the foundation from which all advanced cognitive systems can operate.

With your home base well established, investments across the board begin to yield fruit. The user experience and every interaction yield the data and metrics and nurture a culture of testing that drives the virtuous process that sustainable businesses are built on.

Now you can implement AI within an environment that addresses real gaps in measurable ways and with reduced risk of overshooting.

Applications for AI in Video Programming

Are you a newsroom that needs to grow your audience by increasing watch times? While breaking news gets the most views, it’s not always monetizable - is it possible to optimize content programming to surface videos that are relevant and more monetizable? If so, can this be done within a fluid high-risk newsroom environment where trust is key?

It is not only possible, but it is becoming a requirement to implement video automation and at-scale on a one-to-one level. That path to this is leveraging deep context data gathered from your videos and learning from your audience and how they’re engaging at all levels of your site, holistically across each video experience, across each site and mobile and OTT environment. And put the controls in place to boost only those videos that drive revenue not cost.

Are you a national news network of many sites each in local areas with smaller audiences trying to figure out how to best manage content, ad sales, across each yet tied back to a national presence?

Are you a global sports brand that needs to tie content together contextually, in a deeper way than on top-level category, to package up premium brand-safe and brand-aligned content, in order to realize premium CPM and budget from one of the most finicky ad brands?

Are you a Broadcaster launching a new OTT service, in one of the most competitive environments to-date, with limited information about new subscribers and a need to quickly solve for the “cold start” problem?

Are you a broadcaster looking to prove/disprove the hypothesis that longer content will bring and engage more millennials and Generation Z audiences, prior to making the significantly increased investment in content production?

If any of these scenarios describe familiar challenges heading into 2019 and beyond, then integrating AI into your programming stack and a culture of testing into your organization is your path to success.

A is for Augment

Sabermetrics and Moneyball did not replace baseball GMs, managers, and scouts but rather augmented their intelligence and decades of wisdom and instinct. They are empowered with data-driven insight to make informed decisions. The same is true for a disrupted media and entertainment industry where the stakes couldn't be higher.

For as long as it will be humans consuming content, (I’d like to believe) that it will be important for humans to have a hand in creating and distributing that content. Which means humans and machines working optimally together.

My view on this may change should the world that futurists predict come to pass – where all travel is through automated intelligent vehicles, opportunities for work are limited and there is no longer need for critical thinking, discernment, or free will (should we have ever had it in the first place). I’ll just take my set of personalized whole-life-vitamins spoon fed to give me the life I’m supposed to need.

But we’re not there yet. So let’s celebrate this moment and take control of the content we want to consume and the businesses and society we want to build.