In this case, “shopping” doesn’t include purchasing; however, digital goods and product re-ordering are also poised for growth.

There’s contradictory evidence surrounding consumer adoption of smart speakers for search and commerce. Multiple surveys indicate growth but other data suggest smart speakers have not emerged as a commerce platform. However, eMarketer’s new smart speaker shopping forecast paints a very bullish picture of the future of “voice shopping.”

31 million voice shoppers. The firm says that 31 million Americans “will shop via a smart speaker this year, up 31.6% from 2018.” The company defines “shopping” expansively as “browsing, researching products and adding things to a shopping cart.” In two years, eMarketer says the number will climb to 38 million voice shoppers, as smart speaker owners do more product research on the virtual assistant devices.

EMarketer also says that 21 million people in the U.S. will make a purchase this year using a smart speaker. The important caveat is that these will almost entirely be digital goods: movies and music.

Shopping does not equal buying. There are now a considerable number of consumer surveys in the market that include “shopping” as a question. If shopping is broadly defined, as eMarketer has done, its forecast is plausible, given that there are more than 100 million smart speakers in U.S. homes. Major U.S. retailers also believe that voice shopping will be a major phenomenon within the next three years. However, if “shopping” includes transactions there’s reason for caution.

One category poised for potential growth on smart speakers, however, is reordering. Walmart has built a voice strategy for existing customers around ordering previously purchased groceries. (Kantar previously said that 12.6 percent of Google Home owners have ordered groceries through these devices.) Another potential growth category is “to-go” ordering or food delivery from major restaurant franchises (e.g., Domino’s). This is another version of reordering.

Consumer behavior may tell a different story. There are now so many surveys, you can essentially pick one to support the story you want to tell about voice. A 2018 report from Voicebot.ai, for example, found that roughly 12% of smart speaker owners used voice shopping monthly. In mid-2018 the number was up to 16%. Again, there’s reason to be skeptical. (Smartphones merit a separate discussion around voice search and shopping.)

Yet a recent story from The Information suggests that actual consumer behavior is at odds with consumer survey responses. The story cites Uber as representative of a larger more bearish trend: “Uber receives orders for only a few hundred rides a day through Alexa, said a person briefed on the matter. That amounts to less than 0.002% of the more than 15 million total rides Uber provides per day.”

Why we should care. Notwithstanding the sale of millions of smart speakers, and bullish forecasts, voice shopping and commerce have not really materialized as previously predicted. This is not to say that they won’t or that voice search on smartphones isn’t gaining real momentum.

The central problem in this discussion is that we don’t have analytics or significant behavioral data. We have only surveys, which may not be worthy of extrapolation to the larger population. The smart speaker platforms themselves also need to make usability improvements (and deal with privacy concerns) to realize their full potential. Ultimately, virtual assistants will drive meaningful search volume, including product research and discovery. However, that day may be farther away than expected.

For the time being marketers should pursue local and mobile SEO best practices (there are some specifics with Alexa and Google Assistant), test and experiment (with Skills and voice actions) — preparing for the day when consumers are shopping and buying in meaningful numbers though Alexa and Google Home.