When home goods retailer Williams-Sonoma announced last week that it was dropping $112 million in cash to buy Outward, a 3D imaging and augmented reality (AR) firm, to enhance its own mobile offerings, it raised at least one eyebrow.
Longtime retail watcher Sucharita Mulpuru, currently a retail analyst for Forrester Research, went to Twitter to declare that the acquisition was "a pretty silly use of their cash." Her reasoning was that it's a lot of money, it's a "huge chunk of their cash" and "AR isn't going to beat Houzz, Wayfair, Ikea or anyone else," adding that "I've never seen a retailer recast their destiny (or even make more money really) acquiring a tech firm."
Her argument goes beyond the quintessential corporate buy-or-build debate — although she does wonder if the money wouldn't have been better spent hiring 100 AR programmers, which literally is the classic buy-or-build debate. The essence of her argument is that AR, while nice, is close to a gimmick buy at a time when Williams-Sonoma has serious strategic problems.
This brings us to a key mobile issue. When do shiny newer technologies — facial recognition, emotion detection, AR, etc. — meaningfully enhance the mobile shopping experience, and when are they distractions from more difficult (and more important) retail problems? These technologies, whether bought or built, are typically easier to roll out than dealing with thornier supply chain, pricing, convenience and product mix issues.
Mulpuru argues that W-S went down the flashy and easier route, which she contends will be a serious mistake. Mobile AR right now is at "a very early stage and is very kludgy. All it does is give a very rough approximation of how it may look in a particular room. It doesn't accurately render size. It doesn't calibrate the dimensions of the room you're in," she said in an interview. "It doesn't render color that effectively either. If you can't see the size and color accurately, what's the point?"
I have two concerns with that argument. First, some AR efforts are getting pretty good at projecting room dimensions and comparative sizes. Amazon's AR mobile effort, for example, does an impressive job, based on limited testing I did on the first version of the app this month. And even based on Amazon's own comments, that wasn't the result of a lot of expensive Amazon value-add. Amazon was simply using Apple's ARKit, suggesting that any retailer's iOS app could be as good or almost as good.
And that brings us right back to this $112 million investment by Williams-Sonoma. Why acquire a company or invest in its own team of AR developers when viable versions can be licensed and repurposed? (Note: Williams-Sonoma declined to be interviewed for this column. It did say it would send a statement addressing our questions, but never did.)
The only logical answer is that it thinks its team can develop future capabilities that will go far beyond what others can do. And an acquisition gives it the right to decline extending those magical capabilities to rivals. That's a lot of confidence in a team's ability to seriously outshine others in the future.
I'm inclined to agree with Mulpuru that it doesn't seem likely. Even assuming that a future AR app will deliver absolute precision, how much of a differentiator will that be compared with Google and Apple versions that all rivals will be able to leverage? Google and Apple developers can be quite good, you know.
And while all of this AR business is going on, what does Mulpuru see as the W-S problems being inadequately addressed?
"They are really weak from a supply chain standpoint. Coffee tables take three weeks to get delivered [and many] bedsheets out of stock," she said in our interview. "They have other issues that they need to be tackling rather than focusing on AR. AR is not going to change their destiny. How about spending that $100 million to improve their supply chain? Even throwing a party for their best customers would have a better ROI."
As for AR, Mulpuru said, "it's a feature that it doesn't hurt to have," but she finds that "photographs from customers who bought the item, in their own homes, is very useful."
The essence of Mulpuru's argument is there are key reasons why shoppers buy specific items (price, functionality, convenience, coolness perception, availability, etc.), and none of them is changed by AR. At best, AR makes it easier to visualize something in a room and to make sure that it fits.
Even that "make sure it fits" is tricky. In a kitchen, the room that is arguably more important to W-S shoppers than any other, appliances often have to fit precisely. The dimensions of a refrigerator, for example, don't tell the whole story. Is there a corner that extends just so slightly — just enough to make it difficult to squeeze in? Manufacturer dimensions assume a smooth side where all measurements are equal, but that's not always the case. In short, a perfect fit in an AR app may not translate to a perfect fit in your kitchen.
Contrasting Amazon's mobile AR strategy with Williams-Sonoma's is illuminating. Amazon's supply chain, pricing and product assortment are legendary. In theory, Amazon would be a perfect example of a retailer that has mastered the strategic basics and could justify toying with AR. And yet it opted to simply use what Apple and Google are offering as AR add-ons and focus on its core issues.
W-S, take note.