25 Sep 18 lesson learned 001

This week, I’d like to share some key insights I’ve gained.

The idea started while I was experimenting with Rive and studying how interactive JSON works. I thought: What if I could leverage this to create a platform that enables more dynamic and engaging Google Ads? At first, I believed the idea was so groundbreaking that it could be the foundation of a unicorn startup, so I hesitated to share it with anyone.

But when I began exploring the core question—“Can interactive ads actually run inside Google Ads?”—I quickly realized it was far more difficult than I had imagined. That’s when I understood that keeping ideas to myself was actually the foolish move.

The 150kb Limitation

Google Ads, as I had previously experienced, mostly focuses on images and videos. Like many advertisers, I found myself stuck in the cycle of performance marketing. While not inherently bad, I came to realize that what I wanted to build—whether banners or scroll-triggered animations—ultimately required interactive advertising.

To my surprise, there is indeed an “interactive ads” category in Google Ads, something I didn’t even know existed despite my years of running digital campaigns. If I didn’t know, surely many others don’t either. That gap presents a big opportunity: building a platform where anyone can easily create interactive ads without deep technical skills.

Currently, Google Ads does provide creative tools that auto-generate ads from images and videos, but truly interactive banners still require custom coding. These must be packaged (images + code) into a single compressed file under 150kb—an international digital advertising standard. This constraint forces creativity but also makes it challenging for non-technical creators. That’s why I’m now building templates that allow advertisers to maximize creativity within this strict file limit.

Finding the First Use Case

I believe this service aligns best with experience-driven industries, which also deliver higher ROAS. My first target will be hotels running Google Ads. While researching, I discovered major players like d-edge.com, which offers integrated ad management across Meta, Google, TripAdvisor, and more—plus full booking services. Their product is so well-built that I’d honestly love to work there myself.

Of course, I don’t have the resources or capital to compete with such large-scale providers. Instead, I plan to target smaller hotels that don’t run large campaigns, offering them a pilot version of my service.

Service Evolution & Prototyping

The first experiment was to see if we could actually launch the type of interactive ads I envisioned within Google Ads. Using hotel room photos, I split them into three layers—pillar, bed, and background—and added a parallax effect triggered by mouse hover. Initially, I allowed full 4-way movement, but it felt messy. Restricting motion to horizontal (left–right) made the effect feel more natural and immersive, aligning better with how people visually explore spaces.

Next, I tested a more ambitious idea: letting users explore hotel rooms in 3D space. By stitching together 3–4 images into a panoramic view with WebGL, I hoped to create an experience similar to Google Street View directly inside a banner. However, Google Ads blocks WebGL, and my attempts resulted only in blank screens.

Interestingly, I did find examples of 3D banners running on Google Ads, but it turns out these were not powered by JavaScript—rather, they used formats like .glb 3D models. This led me to another idea: instead of high-fidelity 3D, why not build low-poly 3D representations of hotel rooms, similar to the look of The Sims? It might even feel more engaging and realistic for users. I’m currently testing this approach.