Why Dentsu Inc. Built a Premium Video Advertising Service Now

Dentsu Inc., Dentsu Digital Inc., and CCI have developed "Premium View In-Stream Video Ads," which exclusively gather "premium inventory" from broadcasters' content and other sources, and began full-scale provision this September ( Press Release ).

Why is premium content now being sought as a placement for digital video ads?

This column introduces the issues with in-stream video advertising, their causes, and the solutions the Dentsu Group is proposing.

<Table of Contents>
▼ Brand Safety Issues with In-Stream Video Ads
▼Ad Fraud Issues in In-Stream Video Advertising
▼ Is it Difficult for "CGM" to Check and Remove Inappropriate Content?
▼What is "Premium View," the Safe and Secure Environment for In-Stream Video Ads?

&nbsp;

Brand Safety Issues with In-Stream Video Ads

One of the most popular ad formats for branding purposes is the "in-stream video ad," which is inserted within the content of video sites. The popularity of this format has grown year by year, and it has come to be expected to play a role akin to TV commercials in internet advertising.

However, alongside this rapid rise in popularity, the problem of ad value erosion ( see previous article ) has become apparent. This refers to the "brand safety" risk where a company's ads appear on video content that is inappropriate for advertising, such as hate speech, pornography, or illegally uploaded material.

Such unintended ad placements can have a negative impact on a brand.

Regarding the content where in-stream videos are currently placed, three issues exist from a brand safety perspective.

Recently, cases where ads appear on ethically problematic content—such as adult, grotesque, or hate-related material—have become increasingly noticeable.

While similar issues existed with banner ads, in-stream video ads are a high-impact format (i.e., strong in recognition power). Consequently, negative impressions are perceived more strongly, making them more likely to lead to significant problems.

When problems occur, they directly damage the value of the advertisement.

This occurs when the content where ads are placed contains illegal elements. Specifically, concerns arise about whether the content is illegally uploaded or lacks proper handling of copyrights and portrait rights.

It is not healthy for branding ads to appear on content that is "illegally uploaded" and has "no copyright or portrait rights" processing whatsoever.

While the criteria here are less clear-cut than ethical or legal violations, the quality of the content where ads appear must still be rigorously scrutinized.

Overly restricting placement locations will inevitably limit ad reach. However, indiscriminately expanding placement too broadly carries the risk of ads appearing on low-quality content that could potentially damage brand value.

Ad Fraud Issues with In-Stream Video Ads

Additionally, videos may be played fraudulently by "bots" designed to generate unfair revenue. This "inflated" view count is part of the problem known as ad fraud.

However, ad placements on unspecified media containing such malicious content often appear highly efficient in terms of performance metrics (reach, CPM, completion rate). Consequently, they are frequently highly valued by advertisers, often leading to increased budgets.

This means advertisers not only waste budget unknowingly but also contribute to a negative ecosystem that fuels ad fraud.

Is it difficult for "CGM" to check and remove inappropriate content?

While current in-stream video advertising faces significant issues, improvement is not straightforward. This is because the vast majority of in-stream video ad placements occur on "CGM" platforms. CGM stands for Consumer Generated Media, referring to media sustained by user participation, such as user-generated content.

In CGM, publishers provide users with a "platform" and "a degree of freedom," and users enjoy the service within the established rules. Publishers naturally set various policies from legal and ethical perspectives, but not all users comply. Moreover, it is difficult for publishers to fully monitor all user-generated content and remove only the inappropriate material.

In CGM, content checks for user submissions often follow a flow like the one below, meaning checks typically occur "after" the content is posted. Consequently, inappropriate content remains visible for a certain period until it is deleted.

Furthermore, as the volume of posts increases on major sites, checking naturally becomes more difficult. In this case, manual visual checks inevitably have limitations, making system-based checks the primary method. Consequently, the accuracy of checks decreases, and the relative amount of content that should not be tolerated also increases.

Recently, various efforts have been made even on CGM sites, with some media companies now checking all user-submitted content in near real-time. However, video sites serving as placements for in-stream video ads handle an enormous volume of submitted content. The reality is that human verification and pre-submission checks remain limited.

On the other hand, mechanisms for low-risk video advertising have always existed. The prime example is "TV commercials." TV content is produced by companies, not users, and all such content has been subject to strict "pre-broadcast" screening.

Because all content is checked "before publication," the risk of brand damage is significantly lower than with current internet advertising.

What is "Premium View," the safe and secure environment for in-stream video ads?

The challenges of brand safety and ad fraud in in-stream video advertising have proven difficult to overcome. Creating a safe and secure advertising environment, similar to the TV commercial model, has been a long-standing issue in Japan's internet advertising space.

To address this, we developed the "Premium View In-Stream Video Ads" mentioned at the beginning.

The key point of this service is that it limits the content where in-stream video ads are delivered to "content that is properly managed by companies." Specifically, it ensures a certain level of quality by adhering to the following two rules:

• Rule 1: The placement must not be CGM (Consumer-Generated Media).
• Rule 2: Content must undergo mandatory review prior to publication (meeting standards for ethics, legality, and quality).

Through partnerships with media companies possessing extensive knowledge and experience in video content production, this initiative maintains strict review standards while simultaneously ensuring the quality of the produced content.

Furthermore, since the beneficiaries of the video content are reputable media companies and premium content providers—not general users, as is common with CGM—this initiative serves as a measure to improve and overcome ad fraud issues, including inflated view counts.

We refer to the carefully selected, high-quality video advertising content distributed through this service as "Premium Content."

Through our "Premium View In-Stream Video Ads," we aim to build an ecosystem where creators of premium content receive fair compensation, fostering the production of even more high-quality content. This approach will enable us to continue providing advertisers with a stable supply of ad inventory.