Expect to hear a lot about context in the coming months and years.
We are just starting to understand the immense opportunities for publishers to marry the speed and power of machines with the ideation and creativity that is uniquely human. From content generation tools that can produce a video in seconds, to personalized reader flows that mirror a Netflix experience, to real-time trend analysis for reporters to identify breaking news stories, there is no doubt the “robotic newsroom” will enhance publishing.
Yet the power of context has remained an afterthought at best. Publishers are sitting on a trove of untapped contextual and first-party data that is not being surfaced, organized, or made actionable for media buyers. And in these uncertain times, with unfavorable news surrounding us and the impending death of the cookie, context is poised to emerge as one of the biggest shifts for publishers and advertisers since the advent of programmatic.
Today, many buyers look at context as a brand safety play – or rather, what they don’t want to be associated with. It’s time we view context as a positive targeting strategy. This will require a sophisticated classification system for all forms of content (articles, video, imagery, audio) to not only categorize but to understand the sentiment conveyed in content so buyers can effectively target without brand safety concerns. When the use of contextual metadata in advertising becomes automated, the marketing ecosystem will be primed for a return to a more sophisticated form of contextual targeting. Sellers will have streamlined workflows that increase the value of their inventory, and buyers will know where to find scaled contextual audiences across media properties.
For the golden age of context to arrive, publishers must adopt an AI-first mindset and embrace the change this will bring. Currently, editors will code and categorize the metadata for content in a fairly manual and inconsistent way. They will take an article or video, upload it to the content management system, and log their own metadata including categories and keywords. This process consists of typical categories (i.e. “Sports”) and may include additional layers of information (i.e. “Basketball” and “LeBron James”). Many editors are diligent and put real effort into the coding, while others do not, depending on a number of factors that aren’t always in their control. Furthermore, there’s even some bias that creeps into manual coding, as we are all humans with varied backgrounds, cultures, and experiences.
While this metadata can be useful internally, manual coding falls flat externally and becomes increasingly complex with new media formats. Video, for example, requires an analysis of sight, sound, motion, and emotion to extract organized value for a brand. This level of complexity can be accomplished today only through a variety of AI tools including object detection, sound analysis, and natural language processing. With automation and machine learning woven more prominently into the newsroom, content owners can standardize their coding and produce valuable, insightful audience datasets. So, what should publishers do now to prepare?
Be The Change
Get comfortable with AI and automation being a part of the newsroom. If nobody in your organization is championing AI applications, at the very least to enhance the value of your data, then it’s time to bring it up, put someone in charge of rolling it out, and begin to be the change that moves things forward.
Treat the Data How You’d Like to be Treated
Publishers in particular can find common ground with their competitors by using common taxonomy used by 3rd party data platforms. Once you’ve standardized and organized the attributes (or data) of your content and audience through the use of AI and machine learning tools, you can unlock its true value. Publishers will find they’re extracting more value from their content just by being clearer about what they have and making it more actionable.
Seek Out a Federated Vendor:
While building in-house tech is always an option, finding a partner with pre-built solutions will be a faster route to success. When it comes to working with an AI partner, it is important to find someone who will let you be you. Most publishers don’t need to change the content they produce or overhaul internal systems, they simply need a partner with a federated approach to help identify the value of their content and enhance what they’re already doing.
The speed and power machines aligned with publisher’s creativity presents a once in a generation opportunity to redefine publishing; to reshape content while enhancing revenue. However, publishers must reacquaint themselves with the power of context and become more sophisticated with how they classify content, making it more organized and actionable for buyers. Content is more important than ever, and nothing will ever change that. But with the right amount of automation and AI to create more sensible, cleaner and standardized data, it is possible to make it better for readers and advertisers alike.