Revolutionizing Marketing Asset Management: The Philips Story
In the bustling environment of a marketing department, thousands of photos await their moment to shine. The challenge? Sifting through this extensive collection to find the right visuals for advertising campaigns. Traditionally, this painstaking process could extend over days or even weeks, leaving marketing teams bogged down in monotonous tasks.
One company that confronted this challenge head-on is Philips, the Dutch health technology giant. Capitalizing on advancements in artificial intelligence (AI) and machine learning, Philips leveraged Google Cloud’s Vertex AI platform. This innovative approach allowed them to refine a staggering collection of 200,000 images into a mere 8,000 on-brand images in just a matter of hours.
The Necessity of Visual Consistency
The reality is that maintaining visual consistency across a diverse product portfolio becomes increasingly complex over time. As Claudia Calori, Philips’ head of marketing excellence in personal health, shares, "We manage a large volume of proprietary assets, many of which historically weren’t fully centralized.” The accumulation of various content types made organization a monumental task for the team.
Proper management of these assets is critical. Inconsistencies can dilute a brand’s message and compromise its integrity, making a systematic solution imperative.
The AI Transformation
Recent reports indicate that companies like Philips are now utilizing AI as a powerful tool to hold a competitive edge. According to a PYMNTS Intelligence report, AI has transformed from a mere buzzword into a solid business facilitator, particularly in the goods and technology sectors. A significant 63% of product leaders report using generative AI for product and service innovation. The ability to accelerate creativity, enhance concept development, and streamline visual design processes is a game-changer.
At Philips, a custom algorithm harnessing computer vision technology was developed. This allows the brand to conduct searches and comparisons of images based purely on their visual attributes. No longer limited by file names, manual tags, or outdated metadata, which often fell short in precision, the AI now sorts images based on content.
How Image Embeddings Work
A pivotal component of this innovative solution is image embeddings—mathematical representations that encapsulate an image’s visual characteristics. These embeddings help the AI recognize patterns such as color, shapes, and textures even when images are cropped, resized, or reformatted.
Claudia Calori explains, "Using image embeddings, the system compares degrees of similarity between images across multiple databases and delivers instant results, even when scanning hundreds of thousands of images at once.” This metadata-free approach is driven entirely by the visual content itself, leading to faster, more accurate sorting.
A New Era of Image Management
Once image embeddings are set, the AI conducts a visual similarity search across Philips’ library. Similar images are grouped, and from each group, the optimal version—whether based on quality or recency—is designated as the “master” image. Outdated visuals and duplicates are automatically culled, trimming Philips’ library from 200,000 to only 8,000 images in a matter of hours.
Previously, reliance on metadata to track image variants was cumbersome. As Calori highlights, “As assets were modified, reformatted, or reused, that metadata could easily become disconnected or outdated.” Transitioning to an AI-driven model helps avoid these pitfalls by streamlining the process and enhancing accuracy, allowing the team to refocus on strategic and creative endeavors.
Managing Compliance and Consistency
With the newfound control over digital assets, Philips also enhances compliance across marketing channels. Revised visual assets are traceable back to their original source, ensuring that only approved images are utilized, significantly mitigating compliance risks.
Calori remarks, “It also allows us to strengthen and ensure compliance, automates what used to be a manual effort, and improves the consistency of the assets we present to our customers and consumers.” This enhancement leads to an elevated customer experience, fostering stronger engagement and trust in the Philips brand.
Supporting Future Innovation
Beyond immediate operational efficiencies, the AI capabilities positioned Philips for future adaptations in marketing content strategies. The new system also opens avenues for quicker A/B testing, accelerates campaign launches, and ensures that marketing teams have access to up-to-date asset collections.
Understanding that marketing plays a critical role in helping individuals manage their well-being, Calori states, “Managing our digital assets effectively supports that mission. It allows us to tell a more consistent, clear story across all touchpoints.”
Philips believes that this innovation marks a significant advancement in their asset management strategy, positioning them not just to address current challenges but also to innovate continuously in the ever-evolving landscape of marketing.
Whether you work in a bustling corporate marketing department or are a business leader keen to understand the implications of AI in asset management, the journey of Philips offers invaluable insights into the transformative power of technology in streamlining operations and enhancing brand integrity.