On Saturday afternoon, Elon Musk posted on X that “all federal workers will shortly obtain an e mail requesting to know what they obtained carried out final week.” He added that noncompliance would end in termination. Inside hours, an e-mail from the Workplace of Personnel Administration, with the topic line “What did you do final week?,” appeared within the in-boxes of hundreds of thousands of federal workers. “Please reply to this e mail with approx. 5 bullets of what you achieved final week and cc your supervisor,” it started. The deadline for this activity was Monday, 11:59 P.M. Jap Time. Proper earlier than Musk took over Twitter, in 2022, he had texted this identical query to Parag Agrawal, the corporate’s C.E.O. on the time. Quickly after, Agrawal was fired. It appeared, for the second, that the whole federal authorities would now be subjected to an identical dismissive scrutiny. However then the plan started to disintegrate.
Later that very same night time, Musk clarified the O.P.M. request in a collection of backtracking posts on X. “The passing grade is actually simply ‘Are you able to ship an e mail with phrases that make any sense in any respect?’ ” he wrote at 10:36 P.M. “To be clear, the bar could be very low right here,” he added, ten minutes later. “An e mail with some bullet factors that make any sense in any respect is appropriate!” The subsequent morning, he implied that perhaps the actual function of the e-mail request was truly simply to uncover fraud: “We consider non-existent individuals or the identities of lifeless individuals are getting used to gather paychecks.”
On the identical time, varied Trump-appointed company and division heads started publicly advising their workers to disregard the request. “The F.B.I., by way of the Workplace of the Director, is in control of all of our assessment processes,” wrote Kash Patel, the newly confirmed F.B.I. director, in an e-mail to his employees. “For now, please pause any responses.” Tulsi Gabbard, the brand new director of Nationwide Intelligence, issued comparable directions to her costs, citing the “inherently delicate and categorized nature of our work.” By Monday morning, workers at many companies, together with the Departments of State, Protection, and Homeland Safety, had additionally been instructed to not reply pending additional steerage. Later that afternoon, O.P.M. introduced that answering the e-mail was voluntary.
From Musk’s perspective, the chaos surrounding this hasty order isn’t that essential. As with lots of the current actions of his Division of Authorities Effectivity, what issues is the sign it sends: Musk needs to be seen as a productiveness Prometheus, bringing Silicon Valley’s move-fast-and-break-things effectiveness to the lumbering operations of the federal authorities. If these efforts falter, it’s simply additional proof of the entrenched nature of the forms he’s battling.
The issue with this heroic mythology, nevertheless, is that it’s based mostly on a defective premise. Musk needs the world to consider that the nimble tech sector has already discovered the keys to knowledge-worker productiveness. However, if this was the case, why did Twitter devolve into chaos quickly after Musk’s takeover, as he launched after which cancelled a number of employee-evaluation schemes earlier than immediately firing half his workforce with out additional rationalization? Because it seems, the core query of that O.P.M. e-mail from Saturday—What are workers truly doing?—is one which Silicon Valley itself has been battling since its early days.
Within the nineteen-forties, a younger scholar named Peter Drucker was invited to check the operations of Basic Motors, then the world’s largest company. It was opportune timing for Drucker as a result of G.M. was tackling a productiveness problem. The corporate knew learn how to handle its vehicle factories. Following the strategy perfected by Henry Ford earlier within the century, it divided the method of constructing a automobile into small steps—say, attaching a steering wheel or winding the wire on a magneto—that might then be assigned to assembly-line staff. Productiveness on this context was simple to handle: measure the tempo at which people accomplished their well-defined duties.
Managing G.M.’s workplaces proved trickier. The brand new class of deskbound staff who populated G.M.’s increasing administrative equipment was juggling difficult portfolios of tasks with quickly shifting calls for. There was no single greatest technique to inform them to go about their each day actions, nor was there an equal of a rising stack of steering wheels or magnetos to point how productive they’d been.
Amid this upheaval, G.M.’s long-celebrated C.E.O., Alfred P. Sloan, Jr., started to articulate a radical new strategy to managing the workplace: give attention to outcomes over execution. Drucker remembers Sloan explaining {that a} profitable supervisor on this new atmosphere “should be completely tolerant and pay no consideration to how a person does his work.” It was higher to as a substitute present clear targets after which, later, examine to see in the event that they have been achieved. Drucker absorbed these concepts and ultimately elaborated on them in his influential 1954 e book, “The Observe of Administration,” with a technique he referred to as “administration by targets” (M.B.O.).
When the brand new digital-technology sector started to take off in Northern California, its managers agreed with Drucker. In 1957, Hewlett-Packard launched a set of company reforms, which got here to be often called the H-P Means, and which have been impressed by Drucker’s concepts. “As the corporate grew and it grew to become evident that we needed to develop new ranges of administration, we utilized our personal idea of management-by-objective,” wrote Invoice Hewlett, in a retrospective about this era. Within the early nineteen-seventies, the thought then took maintain at Intel, when one among its unique leaders, Andy Grove, started to aggressively push his personal model of M.B.O., which he referred to as O.Okay.R., brief for Goals and Key Outcomes. (He complained that his outdated employer, Fairchild Semiconductor, valued experience above all else, which led to an absence of “achievement orientation.”)
In 1975, a Harvard Enterprise College scholar named John Doerr made his technique to Intel for a summer time internship. Whereas there, he attended a seminar taught by Grove, who extolled the significance of O.Okay.R.s. As Doerr remembers, “Grove referred to as it a ‘very, quite simple’ system, realizing ‘simplicity’ was catnip to an viewers of engineers.” 5 years later, after Doerr joined the venture-capital agency Kleiner Perkins, he started spreading the O.Okay.R. gospel all through Silicon Valley. In his 2018 e book, “Measure What Issues,” Doerr writes that an O.Okay.R. begins with figuring out an “goal,” which, true to Drucker’s unique idea, captures “WHAT is to be achieved.” This should then be adopted by a number of key outcomes, which “benchmark and monitor HOW we get to the target.” These must be particular and time-limited, and their completion must be measurable. (It was the addition of those key outcomes that differentiated Grove’s O.Okay.R. from Drucker’s M.B.O.) Doerr gives the instance, taken from the unique seminar taught by Grove, during which the target is to dominate a brand new sector within the microcomputer-component market, and one of many key outcomes is to give you ten new designs for a part. The target is formidable however obscure. The important thing outcomes are concrete and one thing managers can observe. On the planet of O.Okay.R.s, the query of “What did you do final week?” isn’t informal however as a substitute a request for measurable updates on specified milestones.
By his personal accounting, Doerr launched O.Okay.R.s to greater than fifty firms, with probably the most well-known being Google. In 1999, quickly after investing $11.8 million for a twelve-per-cent stake of the younger Web-search startup, Doerr gave a PowerPoint presentation about O.Okay.R.s to a small group of Google workers, together with co-founders Larry Web page and Sergey Brin and future C.E.O. Eric Schmidt, who grew to become “tenacious, insistent, even confrontational of their use of OKRs.” As Google exploded into the technology-startup stratosphere, O.Okay.R.s grew to become de rigueur in Silicon Valley. They’re now embraced to a point by many, if not most, main firms, together with Microsoft, Amazon, Netflix, and Uber.
O.Okay.R.s, in different phrases, have change into the closest factor Silicon Valley has to an official reply to the query of learn how to measure their workers’ productiveness. On paper, this management-by-objective strategy makes good sense. In follow, nevertheless, it hasn’t proved to be the silver bullet that its boosters had initially hoped.
A product supervisor at a significant know-how firm lately instructed me that O.Okay.R.s are simple to recreation. “Think about an engineer is tasked with launching a brand new product,” he stated. “They may launch an incomplete venture and arbitrarily declare it carried out, as a result of that satisfies their first key outcome, after which spend the remainder of their time fixing their very own bugs—together with some that they perhaps deliberately put in there—as this satisfies their second key outcome.” When it comes time for a efficiency assessment, the engineer can then brag about all their “progress” on their milestones. “There could be a truthful quantity of manipulation that manner,” he concluded.
Related complaints are simple to seek out on-line. A Reddit thread from final 12 months titled “What’s your expertise with OKR’s” features a commenter who notes, “We’re compelled to make use of it in my firm, however no one appears to make use of it in actuality.” One other commenter admits that many workers at his firm retroactively label work they have been already doing as key outcomes for made-up O.Okay.R.s: “They have been gonna do this anyway, however it now must be labeled as an OKR.” Another person affords an excellent blunter conclusion: “I’ve by no means seen them executed properly.”
Points like these should not the results of simply fixable flaws in Drucker’s unique idea. They as a substitute replicate a deeper actuality: assessing knowledge-work productiveness is a wickedly laborious downside. These jobs are inherently ambiguous with ever-shifting arrays of duties and tasks which are tackled idiosyncratically. This freedom isn’t dangerous. It promotes creativity and a way of agility that makes work much less draining than rote industrial labor, and when harnessed appropriately, it could result in spectacular breakthroughs, like Intel’s microprocessors or Google’s search algorithms. However as data organizations develop too giant for his or her efforts to stay oriented across the drive and imaginative and prescient of singular leaders, this identical freedom makes efficient administration a frightening activity. Even Drucker got here to appreciate that he might have been too optimistic concerning the potential of M.B.O. to completely clear up these points, writing in a 1999 paper, printed greater than 4 many years after he launched his idea, that “work on the productiveness of the data employee has barely begun.”
All of which brings us again to Musk’s performative gesture of asking hundreds of thousands of federal workers to report what they’d achieved through the previous week. Musk would love his followers and followers to consider that this query must be simple to reply, and that he’s reintroducing commonsense productiveness to an obstinate public sector. However the actuality is that, even within the cutting-edge tech world the place Musk earned his unique fortune, this easy question has created limitless complications, together with for Musk himself. When he acquired Twitter, he changed its extra formal O.Okay.R. framework with a extra casual system by way of which he might imperiously and unpredictably demand proof of progress at any time when the temper struck, corresponding to when he instructed all of his engineers to “print out 50 pages of code you’ve carried out within the final 30 days.” However, in keeping with reporting from The Verge, he quickly misplaced curiosity in attempting to evaluate particular person workers, cancelling the threatened code assessment. He quickly started issuing extra erratic declarations, threatening to fireplace anybody who missed a deadline and demanding that managers rank their employees. Engineers have been decreased to monitoring Musk’s Twitter feed to study adjustments to their product plans. Musk’s antics at Twitter is likely to be excessive however it’s true extra usually that, after greater than seventy years of attempting, among the most distinguished firms in Silicon Valley proceed to battle to determine learn how to greatest measure their staff’ productiveness. This can be a main downside worthy of great investigation. But when Sloan, Drucker, Hewlett, Grove, and Doerr couldn’t totally clear up it, a thirty-seven-word e-mail is unlikely to do the trick. ♦
