‘We knew third-party data was useless’: how Stylist got ahead of Google's cookie cull

The move came as a surprise to many, and is reported to have reduced some ad revenues generated on the browser by as much as a fifth. With Chrome accounting for around 66% of web browsing, Google's decision will have a considerably larger impact to third-party dependent revenue.

Google has said that it will consult with the industry and publishers on alternatives to the third-party cookie. Regardless of the solution it proposes, David Hayter, head of digital at The Stylist Group, was unwilling to take a wait-and-see approach: "There’s never been a Google solution that didn’t naturally benefit Google. The fact that most people are signed into their browser probably be its solution.”

Typically, publishers build specific audience profiles at the request of advertisers using first-party data. They identify readers and categorise what they like or engage with to serve (hopefully) relevant ads. But to scale these offerings, third-party data and inferred/lookalike audiences can be used. But Stylist stopped using this kind of data two years ago after being unconvinced by the results.

“Our tests found that the data was useless. It didn't add any value,” he said. The databases were of too low a quality, with Hayter’s own profile showing conflicting attributes. “I was labelled as both male and female, variously aged, unemployed, but earning $150,000 a year. I had no kids and two kids,” he joked. “The point is, most of these were wrong and other publishers we were talking to were seeing exactly the same results.”

Across the industry, in lieu of the targeting that was once on offer, there's also been talk of a return to contextual advertising models. But Hayter's experiences with this approach doen't match the effectiveness of behavioural advertising.

“We looked at some of the results from a contextual campaign we ran and they were terrible. It would been awful for us to rely only on context. I don't see it as the solution,” he said.

Instead, Stylist has assembled around 100,000 actionable reader profiles. Its recent partnership with publisher ad network the Ozone project offers its audiences, along with News UK, The Guardian, The Telegraph and Reach's to reach a supposed "99% of online Brits". Such alliances, once inconceivable, offer larger audience samples in a single transaction, rather than fragmented buys across competing media. The publishers could also theoretically combine learnings and build more effective products.

More advertisers might consider Stylist when it is buying into the niche female audience as part of a broader campaign.

Hayter said: “Ozone now boasts many UK publishers, it feels like there is increasing momentum. We work together although we are to some extent, still competitors. We’ve a much bigger competitor in Google and Facebook who will be able to continue to target in the way that they do at the moment.”

The results of this two year shift in strategy have been promising. Between 2018 and 2019, Stylist Group said it grew digital revenue by 72%.

Forging ahead without shuttered Shortlist, new lead brand Stylist hasn't had the liberty of complacency in carving out a stronger business model.

Stylist remains a freely distributed mag. While full of praise of the “unicorn” magazine brand, Hayter does not believe print solves the problems arising from Chrome's impending update. “One of the things that people liked about digital was the ability to target and proper attribution. Print is not going to solve those problems.”

He concluded: “We worry that fewer people will appraise print and gravitate even further towards Google and Facebook."

To mitigate this, other publishers are rushing to build solutions while there is still time to fill the void.

BuzzFeed has been telling brand partners exactly what their customers are reading across its platforms. With the help of Permutive, it is packaging up clearer audience data for advertisers as its original cookie-dependent data management platform (DMP) was “missing a lot” of audience insight. Invisible to the DMP was activity on Facebook Instant Articles and Google's AMP.

BuzzFeed said it has now doubled the size of the audiences it can sell to advertisers. “We had increased scale, increased visibility, and better audience data all around,” Josh Peters, BuzzFeed’s director of data partnerships said last year.

Meanwhile, UK startup Rezonence is looking to generate new datasets rather than emulate old ones. It believes multiple-choice polls delivered to newsreaders in ad units are a potential solution. Its FreeWall serves polls once a week to news audiences to generate first-party reader data and create email relationships with brands. Poll examples include questions from Pantene about hair volume; Ben & Jerry’s about favourite non-dairy ice cream flavours; Sky Sports about favourite sports; and Nespresso about people's coffee preferences.

Rezonence works with, ESI Media, Hearst, Immediate Media and Bauer Media and claimed to have generated 50m engagements so far.

Consumers that engage are then retargeted with relevant ads and emailed offers and updates. Each engagement is valued at worth around 20p for the publisher, with Rowly Bourne, who co-founded Rezonence in 2014, branding it a clean way to build audience data. "Our users positively opt-in and are retargeted. They know who the brand is, what they want the data for, how they will use it, and for how long," he said.

While there have been complaints about the pricing of quality data, Bourne believes that good data will is due a reappraisal now the “smoke and mirrors” around third-party data have been dismantled.