NRF 2018 preview - retail demos will show if AI and augmented reality are ready for prime time

Jon Reed Profile picture for user jreed January 10, 2018
Not all retail demos are created equal. Here's a couple of recent faves that show how AI and augmented reality can make the retail experience better - and help retailers compete.

With NRF 2018 aka "The Big Show" right around the corner, two things are guaranteed: getting plastered by a blast of cold wind on the streets of Manhattan, and loads of retail demos.

These demos will invariably show off next-gen technology, with everything from "AI" to virtual reality to IoT - anything that might conceivably help retailers in their quest to delight customers rather than lose them to Amazon Prime addictions.

I always leave time to give a few demos a spin. Sometimes I try to charm the demo-giver into shooting an informal show floor video. There are really two types of show floor demos: the futuristic and the practical. I lean towards the practical. If I see two augmented reality demos, I'll film the one that is actually in use by customers. Then we can talk ROI and results - and avoid next-gen pseudo-nirvana.

I'll show you the type of stuff I'll be looking for next week. At the last NRF show, in Los Angeles, I filmed two retail demos that showed how nifty tech can give practical gains. One showed how augmented reality can directly impact online shopping. The other showed "AI" in action for better virtual clothes sizing.

Can augmented reality change the shopping experience? Augment says yes

In How augmented reality is changing the shopping experience - an informal demo with Augment, Jean-Francois Chianetta of Augment gave me the guided tour. Their goal: bring a "physical presence" to online shopping. One key to customer adoption: no VR goggles needed. You can get the AR experience via a mobile device, in this case a tablet:

As you can see at the 1:24 mark, the breakthrough here is the consumer holding a special card behind the tablet, allowing them to see the actual simulated look and feel of the shopping item.

But as Chianetta told me, the goal here is to address a real purchasing problem: mobile conversion rates.

One of the big problems today in shopping is that a lot of the traffic is going to mobile, and the conversion rate in mobile is not as high as desktop.

Enter augmented reality:

We solve that problem by allowing the customer to understand the product by placing it at home.

Augment provides retailers a platform they can integrate with their online commerce. Once products are placed on the Augment platform, they become available in augmented reality. Chianetta says this works for products big and small:

It's integrated in its real size, right in the real environment, so that you can understand exactly how the product is going to fit. It works for big stuff like furniture, but it works also very well for smaller objects, like electronics.

The demo features us kicking tires on a small speaker, which we are able to "hold" to see its contours. The app is called Augment, and it's available on iOS and Android. As of August 2017, it had been downloaded more than three million times.

So the app has been downloaded a ton, but is it working at scale? Chianetta told me their solution is in use by one of Europe's biggest retailers. They're still making in roads in the U.S., but Coca-Cola's sales teams have been using it.

Gut check: does it actually improve conversion rates? We didn't have time to get deep into the results, but yes, Augment's customers have seen improvements in conversions, return rates, and time spent on web pages.

Using AI for... body measurement? A Bold Metrics Demo review

At, a wired-up manequin caught my eye. I soon learned this androgynous model was nicknamed "Manny." Soon I was getting a demo from Morgan Linton of Bold Metrics on how they ease the pain of body measurement and the hassles of online clothes shopping. It's about predicting body measurements and connecting that to garment specifications. Linton:

We've built a technology that very accurately predicts somebody's body measurements... The tape measure shows actually body measurements that we output. We output over 90 body measurements and this is a sampling of them. And what we can do is we can take those body measurement outputs, and we can predict these very easily.

He used Manny to show me how this works:

AI for body measurements and fine clothes shopping with Bold Metrics

But how does the predictive part work? Linton:

We ask people simple questions like height, weight, age, shoe size. We predict over 90 body measurements. We can map those to garments specifications, so brands can understand them.

Here's where it gets interesting: just using a handful of body metrics, they can get within 3 percent of the measurements you would get if you went to see a master tailor:

Think of it like an algorithmic body scale.

Without tools like this, the online shopper's home measurement options were limited to getting your friend to take a picture of you in a tight-fitting body suit, or taking awkward smart phone pictures of yourself:

There's too much friction in that process. We want to make it really easy, so a customer online can get a size recommendation, a brand can get their body data, and it can streamline the entire process.

Manny had a tape measure around its neck. But doing a neck measurement at home is a pain, right?

That's the beauty of it. You tell us your height, your weight, your jean waist, your age, your shoe size, and we predict everything from neck circumference, bicep circumference, thigh circumference. And that data can really help inform brands to understand, "Well, are garments fitting the way they expect?" Because if they aren't, and they're just making guesses about who their customer, is they can change that to make something that actually better fits their population.

Looks cool on the show floor, but has it proven its merit in the real world?

Yeah, we have some of the biggest names in retail use us, and also some of the most innovative, because they're looking out how can they change their process to understand, "Who's buying our products, and how can we make better products for them?"

Off camera, I talked with Linton about some of the trial and error that went into getting the algorithm right. Bold Metrics isn't just aiming to improve the customer experience. They believe better predictive measurements can help companies optimize their supply chains and "make clothes that actually fit the people that are buying them."

Though the holiday shopping numbers were pretty good, retail is a hardcore environment with little margin for error against Amazon's market share and user experience. Next week, I'll be looking for more retail winners, and putting retail tech to the demo test. I hope to see some of you there. Watch this space...

A grey colored placeholder image