Exclusive Q&A: Zenni sees bright future for AI in DTC retail

6/15/2023

Zenni is leveraging AI across its enterprise.

Direct-to-consumer eyewear retailer Zenni is actively adopting advanced AI technology to optimize its digital shopping experience.

Chain Store Age recently spoke with David Ting, CTO of Zenni, to get his thoughts on the rapidly evolving role of AI technology in the direct-to-consumer (DTC) retail vertical. Ting discussed how AI, including generative AI, is transforming e-commerce operations both at Zenni and across the DTC industry in general.

(Editor’s note: Based on machine learning, generative AI can create new content and ideas, including conversations, stories, images, videos, and music.)

What type of AI solutions has Zenni implemented?

When I joined Zenni (in June 2022), I believed that we needed to move to become a data company so we could understand the consumer better. What do they want, their biometrics so we could fit the glasses better for them, their prescriptions. We have implemented a lot of AI techniques and statistical techniques to understand the main eyewear trends, what people need, what people are wearing, and how to prevent customers from making mistakes.

For example, one part our checkout process is a webform. Traditionally, customers fill it out by hand. Zenni found that 7% of customers typed in the wrong information, even though we have over 300 rules in there to prevent errors.

So, Zenni actually introduced an optical character recognition (OCR) solution where customers can just scan the prescription. No matter how sloppy the writing is war, or if the form fields are named differently than on our webform, or if it’s in a different language, our AI is smart enough to input it correctly.

At this point, AI outperforms a human customer service agent in preventing webform errors. The return rate has dropped because 20% of customers are using the prescription OCR solution.

The other AI solution that we introduced is visual search. Zenni built our own set of AI capabilities. Somebody can just take a picture with their phone or a webcam, or upload a picture they find on the Internet, and do that instead of typing words in our search results.

What business issues can AI solve for a direct-to-consumer (DTC) consumer retailer?

AI is going to help build the best consumer experience that makes it seamless and takes away clicks. Prescription OCR is a success for us because it has removed tens of clicks and the stress of trying to put the right number in the right field. And doctor's handwriting may not be the most legible all the time.

The other aspect of what AI is going to help DTC retailers with is properly sizing of the company. Part of Zenni’s mission is to price products at a very thin margin. I'm starting to ask a question that applies before we hired for a position? Do we really need to have this position? Can the AI do this job? If so, let's hold it. Let's use an AI solution and learn how to use the tool.

After we know if AI can do the job or not, then we decided if we're going to hire. We also AI as a way to train up the team on using a tool as quickly as possible. I believe that AI can improve the productivity of each human. It will not replace humans in the next five to 10 years, but makes humans more productive.

When a DTC retailer is able to adopt AI, it should boost the productivity of each employee by five to 10 times.

How will AI transform the DTC space?

DTC retailers could leverage AI to use imagery and also visualization techniques. Online retailers want to digitally simulate the brick-and-mortar experience as much as possible, and with generative AI, can do things like create 3D product views that actually shine in different light conditions.

There's a lot of advancement with AI that will allow consumers to see a new type of shopping experience that's basically going to lessen the disadvantage of not having physical products there by making it easier for customers to find what they need.

Does Zenni have any innovative technology plans to share?

One scenario we are working toward is a customer could pick up their phone in front of their desktop, and then can get their eyes checked. It would take a couple of seconds and the customers could then order a pair because the AI system knows their biometrics from a scan.

Zenni could then give the shopper a perfect pair of eyewear, with the cost roughly one-tenth of what they would pay in a physical store, in less than 10 minutes.

Also, everybody's face is not exactly symmetrical. We are already working on the ability to 3D print an exact frame, and make a mass market of exactly individualized, frames. I think that maybe 10 years from now, fitting for frames will be very individualized because the cost will have come down.

Related Topics