How the Wall Street Journal increases the value of their first-party data with AI

With the continuing rise of consumer privacy concerns and regulations, developing a robust first-party data set has never been more important.

Still, transitioning from a strategy that relies heavily on third party-data can be a challenging process. First-party data sets are often less extensive than the third-party data sets, resulting in an abrupt signal loss for marketing teams. Your first-party data may include customer names and emails, for example, but do you have details on product affinity?

That’s where Artificial Intelligence (AI) can help bridge the gap, by generating predictive insights—such as likelihood to churn—from your first-party data. Those insights enrich customer profiles and increase the value of your first-party data set as a whole.

Here at mParticle, we’ve seen engagement grow significantly when clients combine first-party data with AI insights. On the whole, companies that leverage AI for personalization see:

• 10–20% more engagement (Boston Consulting Group)
• 5–10% higher sales (McKinsey)

The Wall Street Journal (WSJ) is one brand using predictive AI to turn their first-party into a significant competitive advantage.

We recently had the opportunity to speak with Caroline Albanese, Product Director of Data Products and Capabilities for the WSJ, about how her team is generating predictive insights with mParticle Cortex and using them to increase advertising performance. You can check out the full conversation here.

The Wall Street Journal is one of the largest and most respected news publications in the world. Focusing primarily on financial and business news, the WSJ has developed an audience of high-net worth political leaders, law makers, and business executives—all of whom trust WSJ’s journalism and use it to make decisions in their day-to-day lives. 

With a high-value audience and average circulation of over 3.7 million readers, the WSJ is an attractive partner for advertisers. As Caroline explained, her team’s ability to use data to help advertisers reach this audience is a huge differentiator for the publication.

As Product Director of Data Products and Capabilities, it’s Caroline’s job to ensure their customer data set is:

• Operationalized and activated efficiently
• High-quality
• Used to help advertising clients reach the right audience

Prior to transitioning to an advertising segmentation strategy based on first-party data strategy, the WSJ relied on third-party data acquired from data partners. They built cohorts and segments based on that data and used it to help clients reach their audience.

This methodology made it easy for the team to access large data sets, but is also came with some challenges:

• Uncertain data quality
• Privacy and regulatory compliance challenges

The team couldn’t always trust the third-party data they had access to. They were able to activate large segments, but much of the data was questionable. “How many people could really work at Apple? How many financial advisors are really in the world?” Caroline found herself wondering.

Without transparency into where the data actually came from, such questions were difficult to answer, leading the team to question the quality of their third-party data more broadly.

Furthermore, as privacy and user data regulations were ramping up, this lack of transparency was becoming a potential liability, too. Caroline and team did not always have clarity on how data was collected and whether users has been informed their data was collected and could be used for advertising.

“No one knows our content better than us. No one knows our audience better than us,” Caroline thought. “Can we use the proprietary data we're collecting through all these other avenues to create these products and segments—and build that scale over time?”

Instead of paying for generic, broad data of uncertain origin, the team set out to use their own proprietary first-party to create ad products, build audience segments, and help drive results for their advertisers.

“That's what really kicked off our first-party data strategy,” Caroline said, “which dovetailed well with what was already going on in the industry with privacy regulations and cookies going away, and users just being more aware of how their data is being used online.”

First-party data includes any information customers provide directly to your organization. For the WSJ, that includes demographic data and characteristics customers provide when they subscribe to the newspaper, behavioral data from when they visit the website, engage with content, click on ads, share articles, and more.

With regulations on third-party data coming to the fore, a first-party data strategy is crucial for organizations of all sizes and in all industries. But there are some common challenges brands face when developing that strategy, collecting first-party data, and activating it.

Securing organizational buy-in

Any strategic shift needs organizational buy-in, and collecting and harnessing first-party data for advertising is a big change for companies accustomed to using third-party data. 

To prove the success of a first-party data strategy, Caroline started small—replicating a single high-selling audience segment using first-party data—and then grew from there.

She recommends asking yourself, “What is the smallest thing I can build to have an impact?” and then closely measuring and reporting on that impact. “That transparency results-oriented mindset will help you get more people on board, and then eventually get the entire organization on board.”

Data governance and compliance

Using first-party data in place of third-party data enables you to better control for data governance and compliance with user privacy regulations. That also means it’s your organization’s responsibility to solve for that compliance.

Caroline and team worked closely with both the company’s legal and data governance teams throughout data strategy development to ensure the customer data they collected was compliant.

Unifying data across systems

With data being collected by various departments—and stored in numerous different systems—across the organization, centralizing and unifying first-party data can be a challenge for many companies. For example, you may have customer details stored in your CRM, behavioral data from your website and app, and transactional data from your point-of-sale system. Often, these systems are disconnected, and integrating them requires significant engineering support.

Effective first-party strategy requires unifying all your data in one place and solving for challenges like identity resolution to ensure data quality and accuracy.

Targeting effectively with a smaller data set

As we alluded to before, first-party data sets are most often smaller than the broad third-party datasets companies may be used to. Brands are now facing an abrupt signal loss as they transition to a first-party data strategy. 

That’s a challenge for any company, but for a publisher like the WSJ, it’s a crucial one to solve quickly. Advertisers partner with the WSJ in part due to the quality of their ad targeting. Effective ad targeting means more clicks, more conversions, and higher ROI for advertisers—and the publisher’s ability to deliver quality ad targeting hinges on the quality of the data set behind it.

To make first-party data a viable foundation for their ad targeting, the team needed a way to make their customer profiles more complete. They might know a reader’s contact information, country of residence, and recent articles read, for example, but they didn't have a data point on the reader’s product interests or affinity for luxury goods.

That’s where AI comes in, generating predictions based on the first-party data that is available and helping to fill in those gaps in the customer profile—making for more informed and effective ad targeting.

With user privacy and compliance a priority, Caroline and team worked together with the WSJ’s legal and data governance teams to ensure data hygiene, accuracy, and regulatory compliance.

All the customer data generated by the Wall Street Journal —across every department—lands in a central, unified data set. From there, the company leveraged mParticle Cortex—mParticle’s predictive AI engine—to build predictive audience segments and cohorts for advertising campaigns.

With AI-powered segments, the WSJ can build more effective ad segments using first-party data alone—at scale, optimized for CTR and conversions, and adaptive in real-time.

Use Case: Client works with WSJ to reach an audience of financial advisors

An advertiser may want to target anyone with the job title “financial advisor.” It’s not a difficult characteristic to target, but the audience is inherently limited in scale. Plus, the assumption that people with that title will be likely to buy the advertiser’s product is basically a guesstimate.

With Cortex, Caroline and team can identify users likely to be financial advisers as well as people whose behavior suggests they’re likely to become financial advisors or share similar interests with them.

That allows the WSJ to broaden targeting for their advertisers—while increasing CTRs and conversions. Caroline describes, “You want people who work or are very high ranking in the financial industry, but you don't have to only hit those people—you can also hit people who are very likely to be like them or who have very similar interests.”

As Caroline explained, they see much higher conversions, CTRs, and engagement with AI segmentation based on first-party data. Because, at the end of the day, advertisers want people who are interested in their product. The “financial adviser” title is just a proxy for that group. With Cortex, the WSJ team can identify people likely to be interested in the advertiser’s product based on their actual behavior in real-time.

For a media company like the WSJ, that’s a huge competitive advantage because it helps them compete for ad dollars while generating better results for their ad partners.

The company’s first-party data strategy—and focus on data governance and compliance—is now a market differentiator. 

“Our clients are more likely to spend with us because of it,” Caroline said. “We found that our clients trust us more, and we have a much higher return for folks who are using our products, using our first-party data. It's good for us; it's good for our clients.”

For more details on the WSJ’s first-party data strategy—and how they optimize campaign targeting with AI, watch the full conversation with Caroline.