Little known fact: In the Spring of 332 BC, Jeff Bezos’ great75 uncle Jephesius was drawing up the org chart for what would one day become Amazon. Because he still had to wait about 2500 years for the Internet to become a thing, Jephesius enjoyed half-day Friday everyday, passing much of his time drinking wine and playing the ancient Greek version of hackysack (which, played with stones, was far more painful than today’s version).
One random Tuesday in May, a guy named Aristotle–the TED speaker of those days–talked to a packed town square to promote his new book, Metaphysics. He concluded by speaking of a whole being greater than the sum of its parts, which drew massive applause. But Jephesius, who had been nursing the bottle since morning, was passed out in the alley. He missed the memo and so it was written: Amazon’s whole would only become the sum of its parts, not greater.
Some 75 generations later, Amazon became capitalism’s greatest success story, but that success has come largely at the expense of the women and men who’ve built it. Google “worst places to work” and you won’t get very far down the first results page until you see multiple Amazon references. Until recently, I thought this was just a nod to the demanding pace of scale, but now I believe it’s endemic as a result of poorly designed team building; that Jephesius’ drunken Tuesday 2500 years ago has manifested itself through alarmingly high turnover.
Bad Culture Starts in your Job Description
Last Tuesday morning, I spoke to a recruiter about a role on Amazon’s Customer Innovation team. She had reached out to me and the name seemed interesting enough for me to put aside my skepticism. My optimism lasted until just after lunchtime, when she got back to me after discussing my background with the hiring team:
I just finished the meeting with the HR manager and it went well. The few questions she would like to know: What are your data analysis skills? Do you use them every day? If so, how strong are they? The Amazon Innovation team is requiring strong SQL/data analysis skills.
Nine seconds later, I was not only no longer interested in Amazon; I was annoyed that such a hiring manager actually exists. The irony of my reaction? I’ve been using SQL for years and don’t have a problem with it at all.
Out of curiosity, I looked into the job description. Besides SQL skill and “tracking and analyzing metrics on a weekly basis,” the role sought someone experienced in “thought leadership” and “driving large-scale, cross-organizational change.”
I understand that most people wouldn’t bat an eyelash at these buzzwords we’re used to seeing in tech job descriptions, but this is essentially asking a part to be a whole. Second-order effects are not obvious by nature, and I saw in this description a deep case of split personality disorder, in several ways:
- Thought leadership and large-scale organizational change are long-term, values-based initiatives that aren’t trackable in weekly SQL analyses, which speak to optimizations by nature.
- Data analysis has never driven innovation, which comes from knowing how to systematize and replicate a design thinking process that includes ideating, testing assumptions and throwing things into the wild to get feedback from real people. If SQL is in the world of math, innovation is in the fine arts.
- Most people who are drawn to datasets and SQL are not creative thinkers. A recent survey of nearly 8,000 data analysts found that their two most common traits were “investigative” and “conventional”–in other words, more cooks than chefs. They are great at seeing patterns and correlations but typically struggle to hypothesize creatively on causation.
- Likewise, most creative thinkers are not so skilled with big data or interested in learning SQL. Humans are not data points, which is the primary weakness of economics as an explainer of human behavior.
- Humans who are skilled in both analytics and creativity do exist, but they are busy building companies and unlikely to be interested in any company’s shitty middle-management job.
- It takes someone half a day to learn basic SQL; anything more intensive should be going to specialized data scientists anyway. And anything that can be learned in half a day should not be a critical part of hiring criteria.
But what does this have to do with culture?
A common thread of many “data-driven” company cultures is that they struggle to appropriately value long-term, second-order effects of short-term experiments. Very few people are incentivized to care beyond this quarter’s KPIs. Similarly, evaluations are often done in silos as opposed to taking a higher-level view.
In this case, Amazon is recruiting for perfect workercorns, effectively dog whistling know-it-alls and blurring lines of competency in the process. When everyone has the same profile and spheres of expertise are no longer clear, competition (and frustration) typically ensues. A winning basketball team doesn’t need each player to be both point guard and center.
In this “innovation” team, however, Amazon is looking to build a homogenous team of people who might naturally be a 7-out-of-10 in one area while faking their way to a 6-of-10 in the other. In a bubble, this makes perfect sense. But from a macro perspective, a team comprised of a “9” in creativity & influence (“thought leadership”) who isn’t interested in querying and a “9” in data analysis who can’t think outside the box is going to consistently outperform the team of sixes in both. And with both team members able to openly work to their strengths with clear spheres of expertise, both feel valued and empowered without the underlying need to compete.
As one of the world’s biggest companies both in terms of valuation and number of employees, it is likely that Amazon simply doesn’t care about these second-order effects. After all, the company boasted of receiving more than 1 million employment applications coming in as part of last year’s Amazon Career Day; scores of people seem to crave using ‘ex-Amazon’ as two of the 10 words on their short LinkedIn header as though it’s a meaningful intellectual signal. But with ever-growing demand and constant turnover, a recent New York Times analysis suggested that Amazon will need about 5% of the entire American workforce to apply each year simply to sustain.
For all of Bezos’ notable strengths and incredible achievements, few are in his relatability or EQ. He is an innovator who has built an empire on an army of optimizers, and ironically, the kind of lazy, thoughtless job descriptions like the one I saw ensure that only optimizers continue to fill the talent pipeline while innovators with the human skills to drive change stay out of the mix.
Until someone recognizes this subtle conundrum, it’s unlikely that Amazon will make Aristotle proud and become greater than the sum of its homogeneous, SQL-ey parts.