Web-connected microphones -- the kind that come with in-home voice assistance like Google Home and Amazon Echo -- provide marketers with insights into consumer behavior at home. Think of them as
the missing piece that turns disparate channels such as mobile and desktop into a continuous connected omnichannel chain.
It seems to me that a consumer opts-in to having their searches
recorded by the simple task of purchasing and using the device. But what they don't know, and marketers may not know, is that the devices record ambient noise even before the person using the device
provides a "wake-word," such as "Alexa" or "OK Google." It happens on more types of devices than just in-home voice assistants. I've had this experience several times in the past few weeks with Apple
Siri on my iPhone.
Wired reporter Tim Moynihan wrote about what Google and Amazon do with the recorded
data, but my question relates to what Amazon and Google do with the data from the ambient conversations, "the things you say before '"Alexa'" or "'OK Google" to wake up the device.
Moynihan
explains that Google and Amazon do not store the ambient data over a network -- they only store the data that follows the wake word.
Amazon and Google may not store the ambient signals that
come before the wake words, but they do use the words to serve better content. Do they use the signals or words to target better ads?
Does this cross the line from ad targeting to invasion of
privacy, although the consumer gave his/her consent the moment they set up and started using the device?
Signals like these will help Amazon, which plans to release a screen with Echo
next year, serve increasingly accurate content triggered by voice searches -- sometimes with help from Microsoft Bing.
And for those who still need to see the search results rather than just
hear them, Bloomberg reported last week that Amazon is developing an
Echo-like speaker with a screen.
The new device will have a touchscreen measuring about seven inches, per Bloomberg, citing two people familiar with the matter. It will make it easier to
access content such as weather forecasts, calendar appointments, and news, as well as see multiple results from voice queries.
Amazon plans to announce the new feature in early 2017.