A modern analysis of Jeff Bezos’ first-ever job ad for Amazon, posted 25 years ago today, shows that the ad would rate “very weak” by today’s standards and even offers an early illustration of how tech ended up as such a male-dominated industry.
The ad was posted by Bezos on Aug. 22, 1994, on Usenet, the online public bulletin board and discussion network started in 1980. Amazon confirmed its authenticity last year in a story by CNBC, and it surfaced again Thursday in a tweet by BNN Bloomberg’s Jon Erlichman.
Bezos — and his “well-capitalized Seattle start-up” as he called it in the ad’s title — was looking for “extremely talented” Unix developers at the time. Much like the current mentality at the company, candidates were expected to work smart and fast. Bezos wanted them designing and building large and complex systems “in about one-third the time that most competent people think possible.”
On this day in 1994: Jeff Bezos posts the first job ad for Amazon pic.twitter.com/IE32yd7ZET
— Jon Erlichman (@JonErlichman) August 22, 2019
By many obvious measures, the ad worked. Amazon is a tech giant that employs 653,000 people worldwide, and more than 45,000 in Seattle. There are currently 11,108 listings in the Seattle area alone for open jobs at the company.
But the crafting of such job postings has come a long way, thanks in part to the help of technology, and another Seattle tech startup is in the business of helping companies be better at writing them.
Textio, the builders of an augmented writing platform which helps employers analyze and write job ads, among other things, was founded by former Microsoft executives Kieran Snyder and Jensen Harris in 2014. GeekWire asked Textio to use its AI to analyze the 25-year-old ad, and Snyder — who has a PhD in linguistics — offered her own take on the results.
“Our language reveals what we truly value as a culture, and gets responses from people who value the same,” Snyder said. “It seems like Jeff’s ad worked for him to build exactly the company he wanted to build — not so different from Amazon today,” she said, adding that Amazon uses the phrase “competing” six times more often than the rest of the tech industry, and “extreme” five times more often.
“Both phrases statistically attract more men to apply for open roles,” Snyder said.
Amazon’s diversity and inclusion data shows that, as of Dec., 31, 2018, the company’s global gender breakdown is 58.3 percent men and 41.7 percent women. In manager positions, it’s 73.2 percent men and 26.8 percent women. The company says it is continuously striving for better representation across its various businesses. And it’s far from alone among tech companies in employing more men than women.
Here’s a breakdown showing percentage of women at other leading tech companies:
In the first Amazon job posting, Bezos appeared to channel his own persona into what he was looking for in new hires, saying candidates could expect “talented, motivated, intense and interesting co-workers.”
Those interested were directed to mail a resume and cover letter directly to Bezos at Cadabra, Inc., in Bellevue, Wash. (Amazon changed its name from Cadabra, a reference to “abracadabra,” after people on the phone misheard it as “cadaver.” The home where the company got its start — in the garage, of course — sold earlier this year for $1.5 million.)
But out of a possible high score of 100, Bezos’ ad managed a score of 19, or “very weak,” according to Textio’s rating system. It was dinged for such things as needing more “we” statements, being too short, not using enough exclamations, lacking engaging questions, too many adjectives, sentences that were too short and too much “fixed mindset” language.
Snyder, who spent five months at Amazon before starting Textio, has previously written about the difference between fixed and growth mindset and how that language can change the performance of a job posting. Does a company want to hire employees “who have the intelligence and abilities that we are looking for”? or, in exercising a growth mindset, is it focused on how the company helps employees grow?
Textio’s platform analyzes over half a billion documents to make its predictions, according to the company, but for the purposes of Bezos’ job post, the AI looked most closely at other recent engineering jobs in Washington — of which Textio has analyzed 134,024 recent documents.