A New Approach to Cookie-Less Solutions! What is "Persona Targeting" that Captures Latent Needs?

In this article, Ryuya Mizushi of Dentsu Innovation Initiative, who primarily drives business development and R&D with overseas partner companies, spoke with Ryo Matsumoto of Ogury Japan—a company spearheading "persona targeting" as a new digital communication option for the cookie-less era and promoting its domestic service expansion—to gather useful insights for digital marketing in this new landscape.

<Table of Contents>
▼60% recognize the need for cookie-less adaptation. Yet 40% of them have yet to take action?

▼What is Persona Targeting, which enhances target resolution and enables optimal delivery?

▼The difference from Context Targeting lies in reaching potential audiences

▼Key points for improving persona analysis accuracy and gaining insights?

▼By linking with data held by advertising agencies, click-through rates nearly doubled!
&nbsp;

60% recognize the need for cookie-less solutions. However, 40% of them have yet to implement countermeasures?

Mizu: Over the past few years, major platform providers have expanded restrictions on third-party cookie usage. Google announced the gradual phase-out of third-party cookies for Chrome users starting early 2025. With these trends, awareness of the importance of cookie-less adaptation has significantly increased across the industry. How do you perceive this in actual business settings?

Matsumoto: While awareness of its importance is indeed growing, many companies seem to be struggling with concrete countermeasures. Our survey found that "60% of marketing personnel recognize the necessity of cookie-less approaches, yet 40% of them have not implemented countermeasures." Common reasons cited for the lack of progress include insufficient knowledge about alternative solutions, making it difficult to identify the optimal approach for their company, or hesitation in deciding to implement them.

Mizu: With a wide range of alternatives like context targeting (※1) and data clean rooms (※2), it seems common for companies to struggle with determining which solution best fits their product marketing.

Matsumoto: That's right. In our interactions with client companies, we've noticed that industries where targeting precision is particularly critical, such as durable goods manufacturers and luxury brands, tend to engage in more careful and thorough evaluations.

※1 Context Targeting = A system where AI analyzes keywords, text content, images, etc., on the webpage serving as the ad placement, then delivers ads aligned with the page's context.

&nbsp;

※2 Data Clean Room = A secure data environment that allows access for data integration and analysis without identifying individuals.

&nbsp;

What is Persona Targeting, which enhances target resolution and achieves optimal delivery?

Mizu: Against this backdrop, "persona targeting" is increasingly featured in media as a new marketing approach for the cookie-less era, gaining significant attention. Compared to other countries, how advanced is its adoption in the Japanese market?

Matsumoto: While adoption in Japan still lags behind regions like Europe, we see growing domestic interest daily. Our company offers persona-based targeting services across 22 countries, and our Japanese subsidiary's growth pace is exceptionally rapid compared to other international expansions. Ogury Japan was only established in 2022, yet we've already partnered with nearly 200 media companies. Over 148 brands have run more than 500 campaigns, and we're receiving inquiries from many companies. To meet this demand, our Japanese subsidiary launched Ogury Tokyo Lab, initiating the development of advanced solutions tailored to Japanese client needs.

Mizutani: When entering new markets, understanding the characteristics of clients in each country and localizing to meet their needs is crucial, right? We'll discuss Ogury Tokyo Lab's activities in more detail later, but first, could you explain the mechanism and features of "Persona Targeting" again?

Matsumoto: Persona Targeting is a method that uses various data sets, independent of cookies or personal information, to pre-analyze the interests of target demographics. It creates unique personas and delivers ads to the optimal placement surfaces.

To help you visualize it, let me first show you a sample persona analysis. This is a partial analysis of the "Makeup & Cosmetics Interested Group" as an example. As you can see, it allows us to analyze the areas of interest and the media with high viewing tendencies for the subjects being analyzed.

Mizu: In this example, since the target audience is makeup and cosmetics enthusiasts, it's easy to imagine they have a strong interest in beauty-related topics. However, the analysis also suggests they show high interest in other categories like "Food & Drink," "Pets & Animals," and "Shopping." What kind of data is used behind this analysis?

Matsumoto: Ogury's data analysis utilizes non-personally identifiable "survey data," "ad network data (*3)," and "contextual data." Regarding our proprietary survey data collection, we deploy hundreds of types of survey ads across open web ad slots worldwide, generating over 1.5 billion data points daily. We accumulate pattern data reflecting the latest interests and preferences of users visiting these ad slot pages. This data is continuously updated and, when combined with other ad network data and contextual data, forms a data foundation capable of executing massive persona analysis.

Furthermore, the compatibility between personas and each ad placement is scored using our proprietary design. This enables the automatic selection of the optimal ad placement based on the results of the persona analysis.

※3 Ad network data = Ad request data unrelated to personal information. This refers to statistical data linked to the time of user access and the ad placement, analyzed for insights.

&nbsp;

Mizu: Over 1.5 billion data points is truly massive. So, even in a cookie-less environment, this vast amount of proprietary data, not tied to individual users, allows you to understand which personas show interest in what contexts and delivery surfaces? Building such a data foundation must require significant investment and time, making this service a clear competitive advantage for Ogury.

So, how does the process flow from actual analysis to delivery?

Matsumoto: First, we gather the target specifications. Then, we conduct preliminary analysis using the aforementioned persona data on the dashboard. This pre-analysis allows us to output insights on target interests and highly relevant contextual information based on the provided data within Ogury's dashboard. Following this deep dive into targeting through pre-analysis, we implement the customized persona segments for the campaign into the actual targeting at . We then combine these with rich creative assets for delivery.

The difference from context targeting is reaching the latent audience.

Mizu: By the way, we've mentioned the term "context" quite a bit. Could you also explain the difference compared to ad delivery using context targeting?

Matsumoto: Contextual targeting uses AI to analyze the content (text, images, etc.) on the page where the ad will be displayed. It then matches the ad with a placement that complements the product. For example, an ad for sports equipment or sportswear would be shown on a page featuring sports-related content. Unlike standard targeting ads, it analyzes media-side data rather than user-side data, enabling targeting in a cookie-less manner. While it excels at directly reaching audiences where needs are currently apparent—like "sports enthusiasts = sports-related placements"—it also has limitations. Since it requires selecting "which context matches the campaign," it can be challenging to reach audiences where needs haven't yet surfaced, limiting reach to potential audiences.

Persona targeting, on the other hand, excels at reaching potential audiences. Our extensive proprietary research enables us to understand the multifaceted interests and preferences of users linked to each ad placement. Therefore, when targeting sports enthusiasts, we can analyze context that complements them, including non-sports areas. Consequently, ads for sports-related products can be displayed across a broad spectrum of interests beyond sports, effectively reaching potential audiences.

What are the key points for enhancing the accuracy of persona analysis and gaining insights?

Mizu: I'm starting to understand persona targeting better. Could you share key points for actually conducting persona analysis and any findings that emerge from it?

Matsumoto: Let me share an example from a cosmetics brand. Initially, the brand's targeting criteria focused on the classic "makeup/cosmetics enthusiasts" segment. While persona analysis within this broad category is certainly feasible, as shown in the sample at the beginning, this case challenged us to create a more refined delivery design. We broke it down into three categories where the brand is often categorized: "department store cosmetics," "budget cosmetics," and "Korean cosmetics." We then investigated the characteristics of each persona within these categories. This revealed distinct differences within the same cosmetics-interested group. For example, we identified personas with strong support from family demographics (including children's wear and men's fashion) and personas showing high site browsing tendencies for gifts and events. By digging deeper like this, we could sharply grasp characteristics invisible at the initial "cosmetics-interested group" level, enabling a more granular design.

Furthermore, even for the same products, we frequently observed changes in the behaviors and interests of users within a given persona over time. Therefore, rather than continuously delivering to a fixed target audience, it's crucial to regularly update the personas of target users.

Mizushi: I see. Being able to conduct such high-resolution analysis beforehand seems like it could lead to discovering insights that feed back into communication strategies beyond just ad delivery— , for example.

By linking with data held by advertising agencies, click-through rates doubled!
&nbsp;

Mizu: From an advertising agency planner's perspective, conducting persona analysis during product strategy planning and targeting isn't unusual.

However, with traditional digital media delivery, there was a limitation in how well those analysis results could be fully reflected in the targeting itself. In other words, even after conducting meticulous persona analysis, actual ad delivery required matching predefined demographic categories like age and gender, or interest-based segments on the delivery platform. Since these definitions vary across companies, a discrepancy arose between the personas defined in strategy and those actually reflected in delivery.

Ogury's persona targeting addresses this challenge by enabling the strategic personas to be transferred directly into the actual delivery. This represents a significant new value proposition compared to previous approaches.

Matsumoto: Exactly. At the aforementioned Ogury Tokyo Lab, we are advancing initiatives to develop and provide more precise persona analysis. This involves collaborating with data held by advertising agencies and media reps (wholesalers selling ad space to advertisers and agencies) to leverage these strengths.

Mizutani: Within the Dentsu Group, this initiative is also progressing with integration into multiple databases as its first project. I understand that in a trial project with a durable goods manufacturer client, segments utilizing Dentsu Group data for persona analysis achieved nearly double the performance (CTR) compared to segments using standard analysis alone. That's an impressive result.

Matsumoto: That's right. This was particularly remarkable because even Ogury's standalone performance had already been evaluated as superior to traditional digital media. Incidentally, while this initiative started independently at Ogury Japan, it was highly praised by our global headquarters. Thanks to that, it has now expanded into a global effort involving other countries.

Mizu: In our department, we work with many overseas partner companies based in Japan. This is the first time I've heard of a program originating from a Japanese office being adopted into the global business strategy. Especially in foreign tech companies, global headquarters governance is often strong, frequently limiting activities unique to local offices. So, I think this is a very valuable initiative. It shows just how high the expectations are for this project.

Personas Targeting, which seamlessly transfers personas created from each company's data all the way to ad delivery, still seems to hold untapped potential for marketing applications. I'm even more excited to see its future development.
&nbsp;