Future Commerce Podcast: eCommerce, DTC and Retail Strategy

Deep Fakes for Commerce: A New Era of Personalization for Retail

Future Commerce Podcast: eCommerce, DTC and Retail Strategy

CVS launches "Beauty Mark," its truth-in-advertising campaign; while AI is generating plausibly "real faces." Meanwhile, companies like SuperPersonal are putting customers into model try-on videos. Have "deep fakes" - AI algorithms that map faces and micro-expressions onto stock footage - come to retail? How can they help? How can they hurt?

Show Notes:

Main Takeaways:

  • Brian and Phillip are podcasting live from #Shoptalk2019!

  • Deep fakes are getting a little too real to be comfortable.

  • Personal body mapping for try-on is becoming a reality.

  • Can companies figure out how to keep their data in-house?

Who's Waldo: Can Humans Even Spot Deep Fakes Anymore?

Personalization in 2020: Turning Regular People Into Models:

  • Personalization, especially in retail has become a theme of 2019, and the tech is finally catching up.

  • Phillip says that while most virtual try on applications are not very good, Warby Parker has changed the game.

  • Warby Parker's AR powered virtual try on is so good, it's almost like looking in a mirror, and they are using the same depth map as Apple's facial recognition software for iPhone.

  • Another company that's working to change the virtual try-on experience is SuperPersonal, an AI-powered virtual dressing room experience that would allow retailers to "multiply e-commerce photography to account for different ethnicities, skin-colors, and age-groups, without the need to shoot multiple models".

  • "Personalization in 2020 is the whole website is literally you".

  • Brian makes the point that because of the last 6-8 months of advancements in AI and machine vision, models will not be needed, and will only be required as "aspirational content."

Levi's New Story: From Finished Goods to Customizable Clothing:

How Can Companies Get to Know Their Omnichannel Customers?

  • So because 2019 is the year of clientelling, retailers and brands are having to build relationships with their customers, and they need the data to do it.

  • Phillip points out that the more companies aggregate the data in-house and operationalize it as a tech company, the more they will be able to figure out what works, and what doesn't.

  • During Brian's interview with Chris Homer from thredUP, Chris mentioned that thredUP has a policy of testing internally, and figuring out what works in-house, before bringing in tools to supplement those processes.

  • Companies need to figure out what works best for them and double down on that, and they also need to build real systems to house all the data that is collected, in order to utilize it effectively.

There's so much more to see and experience at Shoptalk2019! Stay tuned for more insights, and highlights from the show! Also, let us know, what was your favorite part of #Shoptalk2019 so far?

Go over to Futurecommerce.fm and give us your feedback! We love to hear from our listeners!

Retail Tech is moving fast and Future Commerce is moving faster.

Next Episodes