According to a report published by The Information based on information it obtained from “two people briefed on the company’s internal figures”, only 2% of the people who own Amazon Alexa-powered devices have taken advantage of their ability to make voice-based purchases.

There are estimated to be some 50m Alexa devices in the wild, meaning that 1m of them have been used to make a voice-based purchase. That might not seem like a horrible start, but a number of analysts have touted voice commerce as one of the next big things.

As BGR’s Andy Meek noted, OC&C Strategy Consultants had predicted that voice commerce will generate a whopping $40bn in sales by 2022. If The Information’s sources are accurate, such a prediction looks tenuous at best.

What’s more, The Information’s sources claim 90% of the people who used their Alexa-powered device to make a voice-based purchase failed to make a second voice-based purchase. If that is true, it raises even bigger questions about the future of voice commerce, especially since shopping is one of the features Amazon highlights when promoting its Echo devices.

For its part, Amazon disputed The Information’s sources. “We do not agree with the numbers represented in the article. Millions of customers use Alexa to shop because it is the most convenient way to capture needs in the moment,” it told BGR.

But The Information isn’t alone in calling into question the popularity voice commerce. A report released earlier this year by Episerver revealed that the majority of people who own voice-based intelligent assistant devices aren’t using them to shop. And a study conducted by researchers at Carnegie Mellon University also found low usage of these devices for shopping.

Is voice commerce doomed, or does it still have a future?

While voice commerce looks promising on paper, numerous observers have pointed out that the experience falls short in common real-world scenarios. For instance, obtaining information about and pricing for products using voice commands is probably far less efficient and desirable than using, say, a mobile app. Even for common items that are purchased on a regular basis – “Alexa, buy paper towels” – many customers might want to double-check prices before committing to a purchase, especially if they’re not brand-loyal.

Another observer, an Amazon seller, pointed out that the Alexa keywords for his products frequently get changed, making it more difficult for customers to place orders by voice.

Issues like this might very well explain why owners of voice-based intelligent assistant devices apparently aren’t being used for shopping. In theory, Amazon and other retailers selling through voice-based devices could address some of these issues, and likely will. But as digital commerce evolves, it’s still not clear just how important voice commerce will actually be.

For many product categories, consumers will likely still prefer to engage in digital window shopping before making a web or mobile-based purchase, and with offerings like Amazon Subscriptions and Amazon Dash Buttons, as well as connected smart appliances, it might very well turn out that voice commerce’s useful life was over before it really ever began.

Read our four part series on the state of voice search, or come to the Festival of Marketing 2018 to hear the latest in marketing on our AI & Innovation stage.