Analytics

Quantifying Discovery

The headline finding from Ooyala’s Q1 2015 Global Video Index report was the explosive growth in mobile video consumption over the past two years, now up to 42% of all online video plays and projected to pass 50% before the end of this year. But equally of note for online video providers and marketers were the report’s findings regarding the impact of content discovery and recommendations on video use across all platforms, both fixed and mobile.

According to the report, content recommendations resulted in actual video plays between 35% and 50% of the time — a figure Ooyala calls the discovery ratio –depending on the type of content.

Among news broadcasters, the discovery ratio ranged from 33% to  44%; among sports broadcasters it ranged from 40% to 53%; and for consumer publishers (i.e. non-broadcast) the ratio ranged from 33% to 58%.

More to the point, the report also calculated the amount of “lift” — discovery-driven plays as a percentage of organic video plays — that each site experienced in March as a result of content recommendations.

In general, the amount of lift from recommendations fell in the range of 6% to 12%, but was significantly higher for some publishers. For sports broadcasters the amount of lift ranged from 8% to 16%, while among consumer publishers it ranged from 13% to 21%.

Ooyala Q1 lift ratio

Ooyala Q1 time lift

The data come only from the use of Ooyala’s own recommendation engine so it’s not certain whether the results would hold across all discovery tools. But other recent data points also speak to the competitive and economic value of effective discovery and recommendation tools.

A new branding study by market research firm iModerate found that consumers have far more positive feelings toward Netflix than toward either Hulu or Amazon Prime Instant Video. Among the attributes consumers associate with Netflix is a strong sense that it is a reliable source for the content they most enjoy watching.

“Asked about the brand’s personality, consumers describe Netflix using strong descriptors chocked full of emotion and personification,” iModerate wrote in a report on the findings. “Netflix’s personality is one consumers know and love, and they’re not afraid to share that with you. Netflix’s exciting and entertainment side takes a back seat to its laid-back, cool nature. Netflix is there for you — generous, helpful and resourceful — but be careful, as its engaging nature can be cruelly additive…

“Tellingly, when consumers talk about the service, they describe ‘watching Netflix,’ rather than watching ‘show/movie’ on Netflix. That is unique to Netflix — that consumers often decide to embark on a Netflix binge and then choose a show.” [sic].

The report doesn’t say so, but that sort of brand identification and loyalty among users must be at least partly a tribute to the effectiveness of Netflix’s recommendation system at identifying and surfacing appropriate content.

In contrast, both Hulu and Amazon Prime, both of which started long after Netflix and have much less data to feed into their discovery algorithms, are seen by consumers as channels for watching specific shows, according to the study, but have yet to become a destination in their own right for being reliably entertained.

Netflix and other online services often regard their discovery and recommendation systems as part of their secret sauce, of course, and are often reluctant to share it with others in the industry.

But the Ooyala and iModerate studies suggest content owners and marketers looking to maximize ad views also share a strong interest in effective recommendations and have a strong incentive to identify and share best practices.