Privacy isn’t the only issue. Early use of facial recognition software for marketing purposes has also proved to be less than infallible. For example, if a 20-year-old woman walked in front of an ad, the screen might mistake her for a 40-year-old man, and play the wrong targeted message.

“At least for the first few years, you’ll be seeing ads that are not serving the right person,” says Jennifer Kent, senior analyst at Parks Associates. “Just gender and age themselves will be able to get you some sort of insight on snack behavior but it’s not personal enough to weed out that annoyance factor. It’s really going to take another layer of information to get to the real value of these advertisements. I would imagine that these first trials will help them tweak the algorithms and figure out what’s important.”

Experts say that companies who want to use this sort of technology will probably need to test it well before rolling it out, as well as potentially pilot the program in a smaller number of locations to work out the kinks.

Beyond targeting individual consumers, the most valuable capabilities of the “smart shelf” might lie in building demographic shopping profiles based on how well different products do in different environments.

“You can see the time of day [sales of] some things pick up, or time of week,” says Kent. “Are people shopping for produce or packaged goods at the beginning of the week or later? Can that determine shipping schedules of when different things are delivered?”

From the article, "Smart Shelves Could Move More Triscuits" by Amy Lee.

Be the first to leave a comment.

Have a comment? Login or create an account to start a discussion.