Do you trust Walmart in your home alone?

During its annual shareholder meeting last week, Walmart made an interesting announcement on a new experiment, aimed at the ever-so-elusive online grocery market. As CNBC.com reported: 

Walmart is going to begin delivering groceries inside shoppers’ homes. Right to their kitchen refrigerators. Starting this fall, nearly 1 million people across three cities — Kansas City, Missouri, Pittsburgh and Vero Beach, Florida — will have access to Walmart’s new in-home delivery option, the retailer announced Friday at its annual shareholders meeting in Bentonville, Arkansas. The company said it will “learn and scale” the option across the U.S. from there, not specifically outlining any further expansion plans.

The announcement goes into detail on the exact execution, and in particular discusses the fact that at the scheduled delivery time, the assigned Walmart worker will use a wearable camera device and enter a one-time code on a smart lock to access the customer’s home. There are many things to like, but also many hesitations with this plan. 

Let’s start with the main benefits. While roughly 36.8 percent of U.S. consumers bought groceries online in the past 12 months, just 2.2 percent of U.S. food-and-beverage sales were made online last year.

There are many reasons for this very gradual increase in market adoption of online grocery shopping, including cultural biases, quality control (some people do not trust someone else to pick their produce), but friction is one of the main issues.

While shopping for books or laptops has significantly saved us the time of going to the store while also offering a significant increase in variety, this is not necessarily the case for groceries. When we order groceries online, we usually trade off one inconvenience (going to the store) with another one (waiting for the delivery person).

While this seems to be a minor inconvenience for many, that fact that almost a third have tried it but have decided to use it only very rarely (or never) means that these issues are quite significant and that the time saved is not worth the coordination issues and uncertainty associated with online grocery shopping. 

This brings me to Walmart’s new strategy: If you can reduce the need for coordination between the delivery person and the shopper, you are cutting one crucial (yet, not the only) point of friction. I am quite sure that when Amazon acquired Ring for more than $1 billion, it was about enabling such strategies. 

What are the main drawbacks? I don’t find the camera-based technology to be a viable and scalable long-term solution. The details are not completely clear as to whether the customer needs to watch it in real time (which is clearly a nuisance) or watch it later (which might be too late if something indeed happened). Either approach seems to be a relatively minor reduction in the friction we discussed above.

If the kids left the iPad on the couch, do I need to hide it before I leave home? Is a body camera enough to ensure the person does not “scan” the house for future “endeavors”?

Note that I have no reason to suspect current Walmart employees of any malicious intent. But what about those who know now that having this job means an open invitation to people’s houses, and can Walmart truly alleviate these (baseless?) fears?

I am not even getting into cybersecurity pitfalls and the fact that firms like Walmart have been found in the past to be less than entirely forthcoming about their security breaches. 

The solution may start with the camera, but Walmart will have to go deeper than that. Airbnb has demonstrated that one can build scalable “algorithmic trust,” but it’s clearly far from perfect. 

I am also not sure the solution is very scalable in terms of unit economics. One of the main drivers of the financial success of these online grocery firms is their ability to build density in a given area, which allows the firm to build delivery efficiency (usually measured in “stops per hour”).

The issue is that getting into someone’s house and organizing his or her fridge requires significant time, which is bound to hurt the efficiency of delivery and unit economics.

In a market where customers have shown very high sensitivity to price, there are going to be significant tradeoffs between profit and growth. While growth is clearly a priority, it is not clear for how long it will be sustainable. 

But as Walmart also wrote in the announcement, the goal is to “learn and scale.” As Paul Graham, the founder of Y Combinator, wrote in his legendary blog post, “Do things that don’t scale.”

You need to start with things that don’t scale and find a way to scale them by learning what customers really care about, and you can (or maybe should not) automate them. 

In other words, Walmart’s latest move is a significant development in a market that will eventually be disrupted but clearly takes longer and requires a more sophisticated approach than what we see in the general retail market. 

Gad Allon is a professor of operations, information and decisions at the Wharton School of the University of Pennsylvania and director of the Jerome Fisher Program in Management & Technology.