Traditional audience measurement systems have struggled to keep their arms around all the tentacles of today’s complex, cross-platform TV and video world. In fact, some brands are still measuring audiences using methodologies dating from the era of Mad Men.
At the same time, content owners and pay-TV and over-the-top providers are looking to appeal to their viewers, but also indirectly to advertisers. The secret in their back pocket is that a growing number can now provide brands with granular details about audience behavior — right down to the household and, in many cases, to individual viewers, their viewing behavior, demographics and even their preferred devices.
The audience measurement industry hasn’t rested on its laurels. A number of new analytical technologies have been launched with an eye on this opportunity, including in-home audiometers that listen to your viewing; in-home hardware on the network and in smartphone apps; and tracking from within smart TV and embedded software systems via discovery, recommendations and middleware technologies. These newer technologies greatly enhance the accuracy of the overall picture, which until recently relied on representative audience panels participating in research.
However content owners, providers and brands slice and dice the audience data, though, they must extrapolate the figures, as they are only seeing a sample of the audience at that point in time.
Meanwhile, advertisers are learning the art of patience and now take live audience figures with a pinch of salt, knowing they will rise sharply with catch-up, even for supposedly must-see live events.
Perhaps it is time to assess how the tectonic plates are shifting and examine a completely new approach: One that takes advantage of viewer behavior measurement tools already widely used by pay-TV and OTT providers in promoting content discovery and personalization through the use of AI technologies, i.e., the Netflix effect.
Operators have the infrastructure already in place to measure every single viewer and their engagement with content and channels. Today’s evolving machine learning, AI and big-data tools are transforming operators’ masses of data about consumer behavior into insightful real-time information that can be used for on-the-fly decision-making about content discovery, promotions and advertising.
As forward-thinking operators evolve their focus from counting viewers to using AI to really understand their audience and viewer behavior, they can offer both content producers and advertisers a more compelling proposition.
This analysis has practical applications both internally and externally to influence decisions across marketing, content buying, scheduling and pricing — all in real-time and segmented by different audiences and devices.
Is Anyone Out There?
Instead of simply counting viewers and compiling lists of the most popular programs, providers can answer more challenging questions about audiences and their behavior. These include: Which viewers are most engaged and why? Who is watching, and on what devices at which times? Which content is most popular on which devices, within certain age groups, demographics and geographic regions?
Armed with this greater understanding and granularity, operators can offer a more compelling proposition and explore new revenue-generating opportunities, including targeted ad insertion across multiple screens at the right time and placement.
Look Out for the Netflix Effect
Consumers now have a taste for personalized user interfaces thanks to the pervasiveness of Netflix and the smarter discovery vendors out there who are focused more on AI and getting to know their customers. Operators can use the data to offer viewers a completely personalized experience with recommendations and lists of popular content tailored to their preferences.
Another advantage is that marketing departments can now leave their dusty Excel spreadsheets behind and use new tools to continually evaluate the impact of A/B tests and scenarios, then simply drag and drop assets to fine-tune the mix of editorial promotions and recommendations, the level of personalization and the viewers’ UI.
Audience reach can be optimized by using features including blackout periods, descriptions and rules based on a number of conditions that can be combined such as time zone, age, time of day, language and specific customer segments to target.
The pay-TV industry got very excited about interactive advertising back at the turn of the millennium, but it burned out in the hype cycle. Now, at last, pay-TV and OTT players have a compelling proposition for the advertising industry.
And next time someone suggests installing measurement devices in a subset of homes, remind them that a more sophisticated approach to audience measurement is already in play — one which offers the prospect of additional revenue streams, too.
Peter Docherty is founder and chief technology officer of ThinkAnalytics.
