How Alexa Found Her Voice

Her name was almost “Amazon.”

By “her,” we mean Alexa, the voice-activated assistant within the Amazon Echo.

During the autumn of 2014, there was some disagreement within the walls of the eCommerce giant, as many of the team building what would become the Echo — members of Lab126, Amazon’s hardware division — favored the name “Alexa” for what would become the device’s “wake word” that initiates a user’s ability to give it voice commands.

The most prominent voice of dissent on that matter belonged to a person as influential as they come at the company: Jeff Bezos, Amazon’s CEO.

As Bloomberg explains in a deep dive into the history of the Echo, Bezos’ desire to make the wake word “Amazon” was viewed as problematic by the Lab126 team members by virtue of Amazon’s very success. The company’s name is a word that consumers use a lot, and it’s one that is spoken in frequently running television ads for the electronic retail site. Ergo, the concern was that the device would constantly “hear” its name when it wasn’t actually being spoken to, come online and start attempting to place online orders for whatever it heard subsequently, ostensibly on behalf of its owner.

Could have been a mess.

In fact, as the Bloomberg story details, a lot of steps taken on the path to creating — and selling — the Echo, at times, portended something of a disaster.

Lab126 was created by Amazon in 2004 for the purpose of creating the Kindle eReader. That project — which would come to be known by some members of the lab as Project A (as the outlet explains, the “126” in the lab’s name numerically represents the span of the letters in the English alphabet from “1” (A) to “26” (Z)) — was a success. But that outcome proved not to be the case for Project B — the Fire Phone.

Amazon’s would-be competitor to the iPhone, which launched in the summer of 2014, was selling at a rate not at all equivalent to the phone’s name (unless one is viewing “fire” in the negative sense, like a person’s home burning to the ground), and it was cause for much hand-wringing within Lab126. In the wake of the Fire Phone’s failure, during the final stages of the development of Amazon’s voice-activated home assistant, shares Bloomberg, many members of the team abandoned the project or even their jobs at Amazon.

Work on what would become the Echo had begun, the outlet notes, in 2011. It was known as Project D. (Little is known outside of Amazon about what, exactly, Project C was — or technically is, although work on it has stopped, at least for the time being — other than that the Echo was an offshoot of it.)

In the nascent stages of its development, Amazon knew that it wanted the device to be a speaker. It would play music, and other than that … things were kind of up in the air, according to the Bloomberg story.

Bezos forwarded the notion of the speaker taking on some functionality related to a number of patents that Amazon had filed in 2010 related to “augmented reality” — and, specifically, voice control.

Referring to the CEO’s apparent plethora of ideas surrounding the potential commercial applications for the speaker, a former member of Lab126 told Bloomberg: “There was almost an irrational expectation around the functionality of the device. Jeff had a vision of full integration into every part of the shopping experience.”

Between 2011 and (into the tail end of) 2014, engineers and speech recognition specialists made change after change to the myriad technical specifications of what would become the Echo. However, when one element would appear to be fixed, another problem would pop up in its place and have to be addressed. Tweaks continued to be made, notes Bloomberg, even during what was essentially the final run-up to the Echo’s release.

And when the Amazon Echo was released (first on a limited scale in Nov. 2014), the outlet reminds us, critics came out of the woodwork to throw a great deal of shade on the product, attesting that it was little more than a gimmick that would gain no traction among consumers.

Except, then … it did.

The Bloomberg story shares data from Consumer Intelligence Research Partners released this month showing that more than 3 million Amazon Echos (which launched wide in June 2015) have been sold to date, adding that the device currently holds an average consumer rating on of 4.5 stars out of five.

Beyond the impressive sales figures of a consumer product that was frequently imperiled during its development, one could make the argument (as Bloomberg does) that the more robust promise of the Amazon Echo — and its chances for longevity — lies within the APIs that allow developers to continue adding to and expanding the voice-activated assistant’s capabilities.

In that regard, the future of what the Echo is capable of has yet to be written … and tech innovators from all over the place are jumping at the chance to contribute to that story. is doing its part to help foster that innovation with the Alexa Challenge, whereby 12 lucky companies will have the opportunity to work within the Alexa ecosystem and be mentored by Alexa’s Head of Developer Marketing Paul Cutsinger and CEO Karen Webster during a four-week period (April 27–May 25), following which (on June 15) the PYMNTS community will vote for the innovation that best fulfills the Alexa mission to demonstrate a unique and novel application of voice technology that adds value to consumers and businesses.

All interested innovators (and what innovator isn’t interested in Alexa?) have until April 27, 2016, to submit their ideas via this link.

If your idea is limited to changing Alexa’s name, we would advise against that because Amazon has been down that path before.

And not just with the “wake word,” but with the name of the device itself, too.

As the Bloomberg story notes, not only was Alexa nearly called “Amazon,” the Echo came close to shipping under the name “Flash.”

“Amazon … order my mom a Flash.”

Doesn’t quite have the same ring as what the company went with.



The PYMNTS Cross-Border Merchant Friction Index analyzes the key friction points experienced by consumers browsing, shopping and paying for purchases on international eCommerce sites. PYMNTS examined the checkout processes of 266 B2B and B2C eCommerce sites across 12 industries and operating from locations across Europe and the United States to provide a comprehensive overview of their checkout offerings.

Click to comment