SAP steps into cookie-free future with AI-driven contextual targeting

You are currently viewing SAP steps into cookie-free future with AI-driven contextual targeting


Access versus targeting; Which one will give you better results? This is one of the oldest problems in advertising and is about to become much more urgent for many companies.

The disappearance of the third-party cookie in early 2023 will make targeting online advertising much more difficult. For mass-market businesses focused on reach, this need not be a big deal. But things can be very different for those who sell to specialist audiences. German multinational enterprise software company SAP is one such business.

“As we face the threat of the cookieless future, the challenge of targeting very small but highly specialized audiences will become even more difficult,” explains Moritz Fisecker, EMEA, integrated media specialist at SAP. “We must find solutions first”

This is not the only thing at stake for SAP. The company sees a chance to create a competitive advantage in its industry by preparing for third-party cookie expiration long before it actually happens. He was also looking for ways to make sure his digital advertising was completely brand and brand safe.

Finally, this move is important in terms of the perception of the SAP brand. As Timo Steyer, deputy director of digital strategy at OMD Germany, points out, tech businesses are rarely the first to act on their ads.

“SAP is a business software company, not Coca-Cola, so together we wanted to offer this first mover approach to show ourselves that we’re really ahead of something,” he says. “Anything innovative is worth a try for us. We want to stay ahead of every other big business in the industry and break any barriers to third-party cookies before next year.”

Contextual targeting – nothing personal

There are three main types of solutions for replacing a third-party cookie: first-party data and anonymous identifiers based on that data; interest-based groups, such as Google’s recently announced Topics; and contextual targeting. The second of the three poses the least risk to consumers’ privacy as it does not use any personal data.

Contextual advertising is the oldest form of targeting known to the media. Basically, it is sufficient to place the ads next to the relevant editorial content; If you’re an automaker, you advertise on auto-related pages. The main difference in the modern version is the complexity of matching advertising and content. Until recently, online contextual targeting was based on keyword matching.

Now artificial intelligence (AI) is used to analyze the meaning and responsiveness of a web page’s content, which is then used to choose which ad to serve. In the most innovative solutions, this analysis includes video and audio as well as text.

The result is greater relevance and increased brand relevance – ensuring that not only ads are served next to the wrong content, but also served alongside the right content.

“Brand relevance is about improving the quality of the advertising medium,” Steyer says. “GumGum does this through AI analysis of page semantics as well as video analytics to reduce ad clutter and increase relevance.”

The new kid on the block

SAP had already worked with contextual intelligence company GumGum, Fisecker and Steyer also knew about the programmatic rich media platform JustPremium, which was acquired by GumGum in August 2021. It was Simon Tritsch, DACH commercial director at JustPremium, who suggested that SAP try contextual targeting.

This approach came as SAP rethought its approach to media, as Fisecker explains: “SAP recently celebrated its 50th birthday and we decided to act more like the new kid on the block.”

The JustPremium proposal came in March. Thanks to SAP’s new approach, the project was approved within a few weeks.

“SAP is a big ship, so this kind of quick turnaround wouldn’t have been possible last year,” Steyer says.

The best of both worlds – scale and precision

The product chosen for the trial was a SAP solution tailored for SMEs. The key decision maker in the buying process for these systems is the CIO, which means the target audience for advertising is small.

“They also have a much longer customer journey than an FMCG receiver would have,” Fisecker explains. “Therefore, ads need to be accompanied by ads over a much longer timeframe and across multiple channels. That’s why it’s very important for us to define new ways to find them.”

“GumGum’s new AI targeting means we can combine the best of both worlds; What we need is maximum audience size with sharp targeting,” says Steyer. “Will this hybrid approach mean that we will achieve the necessary awareness generation for both SAP and the product?”

To understand this, experimentation will be measured using traditional brand metrics before and after the campaign.

“We will also measure the pipeline and sales impact,” Steyer says. “We want to see if an awareness campaign can also increase sales. So it’s about relevance, because more relevant ads will drive sales.”

Improved brand relevance will also increase relevance, Steyer explains: “If engagement with content increases, we’ll know we’re doing it right.”

Ultimately, Fisecker says, the benefits of the project are twofold.

“The sooner we start, the more experience we gain. We want to establish a benchmark for future trials,” he says. “But if this experiment works, our new agile media approach means we can allocate more resources.”