Singapore – In a move set to reshape the advertising landscape, at the ESOMAR “Renewal Tour 2025” in Singapore, technology firm MingLue Science and Technology has unveiled AdEff, an AI-powered platform designed to revolutionize global advertising testing and optimization.
Built on MingLue’s proprietary HyperGraph Multi-Modal Large Language Model (HMLLM), AdEff leverages a sophisticated architecture combining large language models and mixed expert models. This enables the platform to swiftly and accurately predict consumer reactions to advertisements, gauge their effectiveness, and provide actionable recommendations for improvement.
Mr. BeiPing Tan, Vice President of MingLue Science and Technology, addressing the audience at the launch event.
According to Mr. BeiPing Tan, Vice President of MingLue Science and Technology, and Head of the AdEff product, “Generative AI is poised to ignite an explosion in marketing productivity, leading to a paradigm shift in the industry. With a deluge of brand content being generated, the ability to swiftly evaluate and select high-quality material will become paramount. AdEff represents a significant breakthrough in the application of global marketing vertical large models for content measurement, signaling the future development of ‘multi-agent collaboration’ in marketing services.”
He further emphasized the platform’s mission: “To provide content measurements that are faster, more cost-effective, and more relevant to brands.” He also announced that AdEff will integrate more brand-generated content measurements going forward.
Starting June 19th, AdEff is available for worldwide trial use. Those interested can register and apply for a trial.
AI-Driven Ad Testing: Accelerating the Selection of Superior Creative Concepts
The quality of advertising creatives is a key determinant of marketing success. A compelling advertisement effectively communicates a unique brand proposition, resonates with consumers, and drives sales growth. Each day, millions of advertisements compete for consumer attention across various platforms. Effectiveness testing before launch is crucial for major ad campaigns to boost reach and completion rates.
Prior to the advent of generative AI, ad testing relied primarily on quantitative and qualitative consumer research and neuroscientific methods like EEG and eye-tracking. These processes often took weeks, involved high costs, and were prone to sample errors.
However, the rapid advancements in generative AI, and especially multi-modal large language models in predicting human responses, have paved the way for a transformative shift in creative testing.
AdEff is a direct result of this technological progress. According to Mr. Tan, it leverages MingLue’s internally developed “HyperGraph Multi-Modal Large Language Model (HMLLM)” in conjunction with mixed expert models. This enables AdEff to simulate consumer reactions to ad creatives and propose tailored optimization strategies within mere minutes. The platform empowers marketing and creative professionals to make swifter, more informed, data-driven decisions, leading to more successful campaigns and a dramatic reduction in testing costs. Businesses can now rigorously test their ad campaigns, striking a balance between creative innovation and business acumen.
Key Technological Advantages: Proprietary Data and Specialized Vertical Models
AdEff’s performance prowess stems from MingLue’s substantial research and development investments in AI large language models.
At its core, the HyperGraph Multi-Modal Large Language Model (HMLLM) integrates non-standard modalities, such as brain wave and eye-tracking data. It utilizes a HyperGraph to bridge the semantic gaps across various modalities, significantly enhancing logical reasoning and understanding in multi-modal contexts. This approach marks a global first in enabling machines to comprehend subjective human experiences.
The formation of the HyperGraph Multi-Modal Large Language Model (HMLLM) is built upon MingLue’s product technology team’s decade-long experience and data in the field of neuroscience, incorporating over 100,000 real-world measurement data points and a vast global creative material library. This foundational data supports the model’s superior performance. The proprietary model has received multiple patents and was nominated for Best Paper at ACM Multimedia (ACMMM) 2024, a leading multimedia conference.
Furthermore, MingLue’s proprietary Mixed-Expert Model (MoE) architecture ensures the precision of AdEff’s predictions. By dynamically scheduling different expert models, AdEff can subjectively comprehend human feelings, objectively analyze the interconnections within multi-modal data, and provide clients with the ability to make rapid, informed decisions, based on expert analysis.
AdEff represents the cutting edge of generative AI and intelligent agents in marketing services, pointing towards the future of marketing tools. During the launch, Mr. Tan announced MingLue’s ongoing commitment to enhancing AdEff across several dimensions.
These enhancements will include expanded content measurement types, encompassing short-form videos through customized models, further technological refinements to enhance the accuracy of AI algorithms and incorporate additional expert assessments to strengthen the Mixed-Expert Model’s analytical capabilities. They will include also personalized adaptation, providing custom model services for specific clients, and globally expanding the ecosystem, offering more language-specific and market-specific versions, in addition to English and Chinese, to meet diverse cultural testing needs, along with enriching creative material libraries to increase predictive accuracy.
As a pioneer in the AI domain with 19 years of industry experience, MingLue is dedicated to fostering co-innovation with worldwide clients and partners to maximize the value of various data sources and unleash the power of AI across multiple marketing scenarios. MingLue’s deep understanding of industry needs allows the firm to provide more intelligent, efficient technology services, driving enhanced efficiency and revenue growth for businesses worldwide.
Visit AdEff.com to embark on your AI-driven advertising testing and optimization journey.
Original article, Author: Tobias. If you wish to reprint this article, please indicate the source:https://aicnbc.com/2928.html