Meredith mines data across its sites for new contextual ad-targeting tool

Meredith is rolling out new ad units with targeting based on a mix of content viewed on Meredith sites and location and time data.

Some examples:

  • A person living on a farm in Michigan reading a story in the morning about how to prepare breakfast quickly, for example, might get an ad for a breakfast sandwich featuring the address of a nearby drive thru.
  • A person living in Miami reading the same article at the same time might get an ad for a breakfast wrap, which the reader can order inside the ad itself, then have the food delivered.
  • A person reading a story about beauty routines in the afternoon might get an ad from a retailer featuring one of the lipsticks mentioned in the article. That same person reading an article about child care in the evening might get an ad from that same retailer offering an accessory that fits one of the strollers highlighted in the piece.

Meredith said the targeting relies on mixing its own first-party data with retailer product information, then matching the right ad with Meredith site content that is most likely to drive a sale or action using the same predictive technology used to recommend content to visitors. If the reader winds up buying something from a retailer that Meredith has an affiliate relationship with, the publisher gets a commission from the sale too.

This past spring, Meredith began testing a prototype of these units on Allrecipes, where it used the tool to sell wines made by the vintner Ste. Michelle Estates. Those ads boasted a clickthrough rate of 5%, De Rubertis said, who declined to share specifics about how much revenue the ads made.

“We’re taking a bunch of classical signals that you’d use with dynamic creative, and doing it with a lot of AI rules,” said Corbin de Rubertis, Meredith’s head of innovation. “It’s about, ‘Can we put something in front of you that is of the maximum level of utility so you’re able to take the next step on your journey?’”

Related
dtc startups
Member Exclusive
‘Flying blind’: How DTC CEOs are preparing for the holidays

Two years ago, Meredith began overhauling the front-and back-ends of its sites, building master user IDs that would allow the company to understand what its audience was reading across its portfolio. That work is still in progress, so these new ad units are only being shown on sites visited by slightly more than half of Meredith’s audience, including Real Simple and People. The rest of the portfolio will be migrated over to the new system by the end of the first half of 2020, said Meredith president and chief digital officer Catherine Levene.

Meredith is gathering its first party and location data from several sources, including consumer data from its 43 million magazine subscribers and millions of Allrecipes readers who consented to share their location data. Eighty percent of the site’s readers consented to share a form of location data so they could see personalized offers from nearby grocers, de Rubertis said; Allrecipes averaged 44 million unique visitors from November 2018 to November 2019, according to Comscore.

Larger publishers are finding ways to better use their first-party user data as typical targeting tools based on cookies fall afoul of new privacy strictures by both browsers and regulators. Vox Media last week announced the launch of Forte, an ad targeting and optimization product built with additional sources of first party data, such as affiliate commerce sales data from its newly acquired New York Media property, The Strategist.

The California Consumer Privacy Act, which went into force last week, is a nice hook, too.

“CCPA is a great marketing tool to make it relevant again,” said Adam Solomon, the CMO of Lotame. “Every [request for proposal] that comes from a marketer has, ‘Here are the demographics, lifestyle attributes, the brands they enjoy, etc.,’” Solomon said. “But no one ever says, ‘I only want people who read this content.’”

https://digiday.com/?p=355620
Digiday Top Stories