Creating Digital Twins for 3D Commerce
3D and AR are proven to increase conversion rates, but how do you get started and do it at scale?
This is a question that many brands and retailers are trying to answer today, and there’s a good reason for it: Augmented Reality (AR) leads to increased sales, reduced returns, higher ad engagements and even lower costs in prototyping.
Early movers in the space like Houzz have seen an 11x increase in conversions for shoppers who have utilized their mobile AR app when shopping. Brands like Ikea have seen their revenue jump by 35% by incorporating 3D assets into their process from design to consumer experience. It is essential for every company to plan a 3D strategy now or they will continue to leave money on the table.
This transition from digital to computer generated images (CGI) is reminiscent of the film to digital transformation that started in the 80s and as we saw then, early adopters reap huge benefits and waiting just increases the pain of switching.
Unfortunately, to date there has been no easy way to create 3D assets for an entire product catalog. Companies who are creating 3D for the web, AR and VR have ended up building custom, one-off 3D models and experiences, which is expensive and time consuming. However, machine learning, image recognition and related technologies are transforming the industry by allowing 3D catalog creation at scale, helping you maximize the potential of 3D across your business. We at VNTANA have been working in the 3D/immersive space for almost a decade, helping brands like Intel, Lexus, Adidas, DJI and others create unique, best-in-class mixed reality experiences. Through this work we have created a robust, scalable content pipeline to solve these 3D model creation issues.
Digital Content Needs
Before we go too much deeper into AR and the possibilities 3D offers retailers, let’s look at the industry as a whole. Last year, eCommerce sales worldwide were $602 Billion and are expected to grow at a rate of 16%.
This is leading to a huge increase in the digital content required to sell all these products. For example, companies like Adidas and Under Armour launch 25K — 35K new products every year and Amazon adds 500,000 new products every day. The best performing products on Amazon have 7 images, so if we do the math, that’s 3.5M images that need to be generated daily and 1.28B images that need to be generated on an annual basis. That’s a ton of content, and that’s just for Amazon.
For this reason, traditional workflows for generating content (i.e. photoshoots & video edits), are no longer feasible for the number of products they are selling both from a cost and time perspective. So, what’s the solution? How do you generate billions of high-quality images & videos annually?
The answer is 3D digital models. 3D models provide:
- more flexibility for customization
- high-quality renders for print and web
- remove the need for expensive prototyping — which also helps improve sustainability initiatives
- millions in saved content creation costs vs lengthy product shoots
3D Models and Organizational Workflows
To create 3D models of products, brands are designing in programs like Clo3D, Browzwear, and Rhino. Or if you design in 2D you can scan your existing products using photogrammetry or other 3D scanning technologies. Either route you choose will have some limitations. For one thing, there is no standard file format output from these programs and the programs are only focused on the designer and manufacturer. Often times, Sales and Marketing departments downstream have to recreate these 3D assets for those channels which is a huge waste of time and money. And the demands on Sales and Marketing are only increasing.
For example, Google is launching Swirl which allows consumers to interact with 3D objects on their ad platform:
Facebook also started testing their AR advertising with a select group of advertisers back in 2018, and has since made this capability available to everyone. Snap has been a leader in AR advertising since its inception. Newer AR advertising platforms are also emerging like those from Unity and Admix. Unfortunately, 3D advertising platforms and big box retailers all have different file format requirements. Below are a few for reference:
Migrating to 3D Models for Product Visualization
So, this brings us back to our original question. How do you compete in a world where consumers are increasingly expecting the ability to interact with your product in a CGI world?
If you design your products in 3D today you are one step ahead. Connecting your design and manufacturing departments with Sales and Marketing to let them know these 3D assets exist is key. By getting better alignment between teams, brands and retailers can begin to invest in software solutions that enable their 3D assets to be used across the entire organization.
Once you have alignment, you will need to create a holistic software plan to ensure original design files will be usable by programs Sales and Marketing are using. Often times design files are huge because they have every detail for manufacturing so they must be optimized. This means decimating the mesh to reduce polygon count, removing obstructed geometry and compressing material files. Then the file must be converted to the file formats for each use case. Below are a few examples:
- USDZ for ARKit
- GLB for ARCore and most web viewers
- Unity Asset Bundles and Packages for mobile games, AR and VR experiences
- Unreal Asset Bundles and Packages for AR, VR and other interactive 3D experiences
VNTANA’s 3D Automation Platform is one solution for optimization and conversion. Our algorithm actually understands the shape of the model and then optimizes it, shrinking the size by up to 90% with virtually no shape distortion or data loss. Then it converts to all of the above file types so you can use the 3D model across all platforms.
Size, poly count and number of vertices are all important factors in where a 3D asset can be used. VNTANA’s software understands the shape of the model first, then creates the minimum number of vertices to create that same shape while removing obstructed geometry, giving you a high quality result with minimum size, poly count and vertices. Depending on your use case there are different requirements as mentioned above.
In order to know what is right for you, first do an audit of all programs currently being used across your organization and be sure to list your file format import and export requirements. This will help you map out your pipeline and find holes where you need a new solution to bridge the gap.
In short, to compete in an AR world you need to get your fundamentals right first. Anyone can build a one-off AR app to try on a specific product or see what it might look like in your home, but the real winners are going to be the companies who have a cohesive, integrated strategy for creating and distributing 3D assets at scale from design to marketing.
