Data-driven Creative Evaluation: AXer Launches System for Quality Assessment of Produced Images

AiTRIGGER Inc. will begin offering a new service resource within AXer in July 2026 that digitizes the quality of produced images (such as eye-catch images), scores them based on designer evaluation criteria, and even suggests areas for improvement. The company has already implemented this system in its own public relations creative production workflow and commenced operations. The initiative stems from the challenge of judging the quality of creative work, which often depends on the subjective senses of specific individuals, making standardization and talent development difficult. This is not a tool implementation, but rather a practical application of Marketing AX, transforming the evaluation process itself with AI.

The Challenge of "Subjective Evaluation" Identified in Public Relations Creative Production

AiTRIGGER has produced a large volume of creative assets for various projects, including advertising banners, landing pages, and content eye-catch images. In this process, the workflow involved new team members creating images, followed by in-house visual inspection.

What became apparent was that the criteria for judging the quality of creative work were dependent on the subjective senses of a few experienced designers and directors. The definition of "good" was not verbalized, and the level of detail in feedback varied depending on the reviewer. Consequently, quality tended to be inconsistent, and it took a long time to train new members.

When evaluation is subjective, increasing production volume leads to bottlenecks in review and rework. To address this challenge, the company began building a system that digitizes the evaluation criteria themselves, ensuring a consistent level of quality regardless of who creates the work.

Transforming the "Evaluation" Process Itself with Standardized Data

*This is a demo screen.* *This is a demo screen.*

AiTRIGGER's approach was not to provide a convenient scoring tool. Instead, the company aimed to transform the evaluation process itself by verbalizing the evaluation perspectives that designers hold internally and embedding them as data. This system brings the following changes to the production workflow:

1 Subjective judgments become shared criteria

The quality judgments, previously left to individual senses, have been organized into evaluation perspectives. Generative AI reads the produced images and scores their quality based on these criteria. Both the creator and third parties can now see what is good and what is lacking in the same terms.

2 Feedback extends beyond指摘 to include directions for improvement

Instead of just providing a score, the system presents both areas for improvement and aspects that are already well-executed. Creators can then independently determine what to revise next.

3 Training and quality standardization progress simultaneously

New members responsible for production can receive feedback based on the same criteria. They can gain experience while maintaining a certain level of quality without constant supervision from a mentor, enabling both faster training and improved production quality.

This system will not be offered as a standalone SaaS tool. Instead, it will be implemented as part of the resources (people, tools, know-how) provided by AXer, tailored to each company's production workflow. For detailed service information, please visit the service page.

Demonstrating the Design to "AX-ify" Evaluation, Not Just a Single Feature

This initiative is not merely about a single image scoring function. It is an attempt to redesign the evaluation process, which has traditionally been confined to individual human senses, into a shareable design. AXer provides resources that design, build, and operate entire business workflows by combining people, tools, and know-how.

The workflow shifts from one where people judge quality based on experience and intuition to one where AI evaluates based on established criteria, and humans make decisions based on that AI evaluation. The company has begun implementation and operation within its own public relations production and will sequentially offer this service to other companies facing similar challenges with subjective evaluations. This system is not a fixed, pre-packaged solution but is reconfigured according to each company's production assets and standards. Consultations can begin from the problem-definition stage.

AXer Service Page https://aitrigger.co.jp/service/axer/

AiTRIGGER Inc. Headquarters: 16F, D Tower Nishishinjuku, 6-11-3 Nishishinjuku, Shinjuku-ku, Tokyo Business Activities: Digital marketing support (performance advertising, in-house support), Marketing Resource Management (MRM), Marketing AX business URL: https://aitrigger.co.jp/