With third-party cookies on their way out, direct deals are making a comeback. For newer publishers that have yet to integrate open-auction infrastructure, there’s no rush to start a programmatic business.
One such publisher is Likewise, which offers personalized TV, movie, book and podcast recommendations.
The company launched in 2018 with backing from Bill Gates, who serves as a strategic advisor. Since its founding, Likewise has focused solely on monetizing through direct-sold ads and sponsored content, rather than programmatic or paid subscriptions.
Likewise has steered clear of programmatic for several reasons, CEO Ian Morris told AdExchanger. Since programmatic is all about advertisers looking for “low-cost impressions,” he said, the company gets more revenue from direct deals that rely on its first-party audience data.
And, he added, programmatic offers little control over what ads appear on the publisher’s pages.
“We’re creating dedicated emails and having fun mixing content and ads,” Morris said. “Our design team is creating interesting things partially because we haven’t standardized everything yet.”
Data points
All that being said, Likewise isn’t ruling out programmatic.
The company has experimented with programmatic display placements, Morris said, and he could see the business eventually using open auctions to backfill inventory it can’t sell directly.
However, native content will remain its main focus for the foreseeable future, he said.
The foundation for these direct deals is Likewise’s first-party data on the media preferences of its six million registered users. When users install the Likewise app and register with their email address, they are prompted to answer a survey on their favorite genres and media properties, which is used for recommendations.
Nearly 90% of Likewise users provide some information on their media preferences on signup, Morris said, and more than 80% provide demographic data, as well. Likewise also collects data on what media its users have added to their watchlists and what they’re recommending to friends.
Though Likewise can easily target specific email addresses, Morris said the company’s approach to ad targeting is a mix of user-based and contextual.
“We’re looking at the audience and saying this person is using us to find romance novels, so let’s not hit them with nonfiction or space opera,” Morris said. “But we’re also looking at where the advertiser wants to get placed.”
Nonendemic native ads
Likewise’s first-party data can be used for straightforward endemic campaigns. For example, a user who shows an interest in horror movies might see takeover ads and paid recommendations for the new horror movie “Thanksgiving,” whereas parents would see ads for “Paw Patrol.”
But the data also opens up native advertising opportunities with nonendemic advertisers. Likewise’s audience is 80% Gen Z or millennial and more than 70% female, which are key demos for brands in the automotive, travel and financial services verticals, Morris said.
That’s allowed Likewise to diversify beyond its initial advertiser list of book publishers and be less dependent on TV and entertainment industry ads, which saw a slowdown when the Hollywood writers and actors strikes froze new releases.
For example, Likewise has worked with travel companies on custom email blasts. One recent email activation aggregated media in which New York City is a central character rather than a setting, Morris said. It featured native ads and a storyline centered on New York City, with hotels and destinations targeted to audiences who have signaled an interest in travel.
Curacity, a platform that helps email newsletters monetize travel-related content with ads for hotels, has been working with Likewise on creating branded travel content, said its head of media brands, Evan Agaston. It also helps Likewise attribute campaign performance by measuring how a hotel’s revenue and guest signups were impacted after a campaign.
Generative AI as first-party data source
Although Likewise already has a well of data on subscribers, it’s also experimenting with ways to target and collect data from non-logged-in audiences. That’s one of the use cases for its recently launched Pix AI chatbot, which was built on OpenAI’s technology, Morris said.
Even if a user hasn’t registered an email address, they can visit the Likewise site and ask Pix for content recommendations. Those interactions are incorporated into the AI’s machine learning data and can be used to model additional insights for targeting logged-in users. For example, if the AI observes a correlation between horror fans and true crime fans, that signal could prompt Likewise to recommend true crime content to logged-in users with a preference for horror.
When subscribers use the AI, they receive more personalized recommendations – including direct links to open media in the streaming apps they have installed. However, at this time, Likewise lacks integrations with these apps that would allow these links to be used for attribution.
Ultimately, Likewise’s investment in AI is also an investment in first-party data, and it complements the company’s mission of understanding what content people want to consume.
Going forward, Likewise is interested in exploring the possibility of sharing first-party data with advertisers for retargeting purposes – like serving an ad for “Thanksgiving” on a third-party site where a cookie match finds a connection to a logged-in user who expressed interest in seeing that movie.
But for now, it’s focused on activating its first-party data in its own ad products, Morris said. “We’re working with advertisers who’ve been at this for a long time, and a lot of the demand to date has been for first-party data.”
