For streaming video apps, data is the only way forward

With thousands of titles and dozens of providers to choose from, viewers need to be aware of the types of content you have available — at best, they will know one or two shows, which is likely not representative of your full catalog.

Netflix has set a very high bar on data use. It has been able to gain a clear lead over the rest of the industry by investing sizable sums into creating and tracking data — by some estimates, as much as $150 million a year.

More important, it has put this data to good use. Netflix utilizes its data for marketing purposes in order to increase its audience, currently estimated at over 50 million subscribers in the U.S. alone. It can target consumers with titles by interest, and it can then use its data to find similar titles as well. Consumers who like “Raiders of the Lost Ark” may also like “Out of Africa,” and by showing them both titles in a targeted ad, Netflix vastly improves its chances of gaining (and retaining) a customer.

Investing in original content is an expensive undertaking, though. Netflix uses its extensive data to eliminate some of the risk. By focusing on the types of shows that its users are likely to respond to, Netflix has been able to make more informed decisions about the types of shows it should invest in. That was the case with “House of Cards,” which became Netflix’s first foray into original programming once it had determined that political dramas would be an excellent fit.

The final area where Netflix makes use of its data is in site personalization. By looking at what shows you’ve watched, it is able to suggest similar shows and movies you might like, creating a unique and personal experience for every user. This has proven to be a very appealing feature, as it generally turns up shows the viewer was familiar with but may not have known were available on Netflix, increasing their appreciation of the breadth and depth of Netflix’s content.

Anyone launching a new so-called over-the-top or streaming site must be prepared to live up to Netflix’s standard, as consumers will now expect that level of personalization.

Since most entrants can’t afford to sink $150 million a year into data, their solution will need to come from third-party vendors like Jinni, which originally specialized in building recommendation engines. By pairing “taste partners,” users who like the same movies, Jinni is able to give new OTT sites the ability to offer greater personalization.

It is also able to use its own databases to help fill in the gaps in its partners’ databases, which then allows providers to begin pushing titles from deep within their catalogs directly to the people who are most likely to be interested in them. Providers also have the ability to surface suggested content as part of an advertising campaign, which serves as a way to informing customers — both current and potential — how much programming is available.

“With personalization, you put the user in the center and ask which titles are a good match for them,” notes Jinni CEO Yosi Glick. “With targeted advertising, you put the title in the center and ask which is the right audience. One challenge is the mirror image of the other.”

As personalization becomes the watchword for streaming video, data becomes more important. While many new entrants are tempted to throw all their money into content acquisition, that strategy will fail if it isn’t supported by a rich data play: The more providers know about their underlying databases, the more use they can make out of the content they have, the more cost-effective that content becomes.