Bottom Line Up Front
The problem. Many workplace targets run on the same rule as the Apple Watch’s Monday Move goal. Hit this quarter’s number and next quarter’s goes up, with no point at which it stops.
Why it happens. Rising targets work well for people starting out. A 2013 walking trial found adaptive goals beat a fixed 10,000-step goal, and progress bars and streaks speed effort up near a reward. Soviet planners ran the same rule on factories, and economists call it the ratchet principle.
The risk. People facing a ratchet hide capacity, as Soviet managers, American machinists and the Hawthorne wiremen did. When they cannot hold the number down, some fake it up, as Wells Fargo employees did with more than 2 million unauthorized accounts. Streak-style rewards discourage honest reporting. Output metrics count only what ships, so a week spent sick, caring for a parent, training new hires or recovering from an outage shows up as a slow week.
What to do. Set targets from what the work requires, look back over a quarter or more, and write down which events lower the bar. Audit how numbers get met, reward reporting over streaks, plan for the dip after a big push, stop improving a number once the next increment costs more than it returns, and decide in advance what level counts as enough, for the team and for yourself.
The Man Who Wanted It to Stop
A few years ago, a man posted a question on Apple’s support forum. By his own description he was a senior citizen who loved his Apple Watch and had worn it for years. He had set his activity goals where he wanted them. He met them, week after week.
Mondays were the problem.
Every Monday morning the watch sends a summary of the previous week and proposes a new Move goal, built from what the wearer burned over the past seven days, according to TechRadar. Because he kept meeting his goal, the watch kept proposing a bigger one, and he wanted the proposals to stop. Switching off Move goal notifications altogether, progress updates included, was the answer offered in the Apple Support Community thread. No one in the thread suggested how to hold the goal steady and keep the reminders.
He was doing, steadily and for years, exactly what the watch was built to encourage, and the system answered his success by asking for more.
“Like everyone, I’m addicted to closing the rings,” Jay Blahnik, who helped design them at Apple, said in a 2015 interview reported by AppleInsider. Behind that design sit ideas tested across most of the last century, on a Soviet factory floor, in a maze full of rats, with a café loyalty card and among a group of lottery winners in Illinois.
The Plant Manager’s Problem
Imagine running a Soviet factory in the 1950s. Moscow sets your production target for the year; hit it, and you and your workers collect bonuses. Now suppose you have a very good year and exceed the plan by twenty percent.
What happens next year?
Planners take what you produced and set the new target from it. Soviet officials called this planning from the achieved level. In a 1980 paper in the Bell Journal of Economics, the American economist Martin Weitzman modeled the incentive problem it creates under the name it already went by, the ratchet principle.
Managers learned fast. Interviewing former Soviet managers for his 1957 book Factory and Manager in the USSR, the economist Joseph Berliner documented how they hoarded plant capacity, labor and inventory and rushed production at the end of each plan period to make the quota. Beat the plan by a little and no more, keep spare capacity out of sight, never let the factory look as good as it could. Salespeople facing quotas built from last year’s number hold back deals the same way and call it sandbagging.
Monday’s Move goal works from the achieved level too, with last week’s output as the reference for this week’s target.
Twenty Walkers
Twenty overweight adults wore pedometers for six months in a 2013 trial run by the public health researcher Marc Adams and colleagues including James Sallis. Half got the standard 10,000 steps a day. For the other half, an algorithm looked each day at the participant’s own recent step counts, ranked them and picked a percentile as the next target, so the goal rose and fell with the walking. By the end, according to the trial abstract indexed by PEDro, the adaptive group had out-walked the 10,000-step group.
Adams called the method percentile shaping, a technique borrowed from operant conditioning, the branch of psychology that taught pigeons to peck in sequence. Someone who walks 3,000 steps on an ordinary day faces a fixed 10,000-step target more than three times their norm, while a goal pegged to their own recent days sits just past their usual day, every day. A later program, described in the Journal of Medical Internet Research, ranked a nine-day window and set the target at a percentile such as the 60th. Hit the 60th percentile often enough and your ordinary day rises, pulling the percentile and the goal up behind it.
From the couch, the method works; the Adams trial showed as much. Apple doesn’t publish the formula behind its Monday suggestion, though the watch behaves the way percentile shaping predicts. Trouble starts later, once the person on the other end has already arrived where they wanted to be. The method has no rule for leaving the goal where it is.
The Last Stamp
Rats in a maze run faster as they near the food. Clark Hull, a Yale psychologist, timed them in the early 1930s and called the effect the goal gradient.
Seventy years later, Ran Kivetz, Oleg Urminsky and Yuhuang Zheng took the idea to a university café, where customers got a card promising a free coffee after ten purchases. As the cards filled, the time between purchases shrank by about 20 percent, per a summary in the Dartmouth Undergraduate Journal of Science. Then came a trick. Some customers got a twelve-stamp card with two stamps already filled in; others got a blank ten-stamp card. Both needed ten purchases. Pre-stamped cards were finished faster.
Right after the free coffee, activity dropped, and members were at their most likely to drift away, the same 2006 paper reported in the Journal of Marketing Research.
Both mechanics are built into the red ring. Filling all day like a stamp card, it shows exactly how much distance is left and resets to empty at midnight. Streaks add a second pull. In 1979, Daniel Kahneman and Amos Tversky showed that losses weigh on people more heavily than gains of the same size, and under that finding a long streak behaves like a possession, so missing one Tuesday registers as a loss.
The Lottery Winners
Winning a large prize in the Illinois State Lottery did not leave people rating themselves meaningfully happier than a comparison group of non-winners, according to a 1978 study by Philip Brickman, Dan Coates and Ronnie Janoff-Bulman. Winners did report getting less pleasure from ordinary things, a joke, a compliment, breakfast. People paralyzed in accidents, interviewed for the same study, rated their present happiness lower than the comparison group but still above the middle of the scale.
Seven years earlier, Brickman and Donald Campbell had described this adaptation as the hedonic treadmill. Each gain becomes the baseline people measure from. A raise turns into the salary within a few paychecks, and a promotion turns into the job.
The Monday algorithm runs on the same rule, and so do most career ladders and quarterly targets.
What the Wrist Knows
Heart rate from an optical sensor, motion from an accelerometer and gyroscope, and the age, sex, height and weight entered at setup go into every estimate. From those inputs the watch calculates calories burned, subtracts what the body would have burned at rest, and calls the remainder active calories. Blahnik described the Move ring to NY1 in 2015 as every movement of the day expressed in that one unit.
Plenty of effort falls outside it. Pushing a stroller for an hour, a parent keeps the wrist nearly still on the handle. Carrying boxes up three flights of stairs produces a brief spike. A nurse on a double shift and a person sitting all afternoon beside a parent’s hospital bed get scored by what the wrist did, nothing more. Sleep doesn’t register. Neither does a fever, or grief.
For nine years, from the watch’s 2015 launch until watchOS 11 in September 2024, a sick day ended a streak, per MacRumors. Apple then added a pause, along with goals that can differ by day of the week. It is a manual switch. Whether its wearer has the flu, the watch still cannot tell; someone has to open the app and tell it.
Seven days is all the Monday suggestion looks back on, while the things that change what a reasonable day of movement looks like, a newborn, a new job, a diagnosis, a winter, a body getting older, change over months or years.
Lessons for Managers
Sprint velocity, sales quotas, annual operating plans and performance ratings often start from what a team did last period. American workplaces have a long record of what that rule produces.
Rising targets teach people to hide capacity
For eleven months in 1944 and 1945, the sociologist Donald Roy ran a radial drill on piecework in a machine shop he called the Geer Company. Operators there held output below what they could produce so management would not cut the rate, and Roy published the account in 1952 as “Quota Restriction and Goldbricking in a Machine Shop” in the American Journal of Sociology. Thirty years later Michael Burawoy took a job in the same plant and wrote Manufacturing Consent (1979) about it. A decade before Roy, researchers at Western Electric’s Hawthorne Works in Cicero, Illinois, watched the men of the bank-wiring room keep output down for fear the company would lower the base rate, according to Wikipedia’s account of the Hawthorne studies.
A dashboard that tracks output against a rising target loses sight of a team’s real capacity. Set targets from what the work requires, and say in writing that a strong quarter will not automatically raise the next one.
When the number can’t be held down, it gets faked up
Wells Fargo admitted to the Justice Department that it increased its focus on sales volume and annual sales growth beginning in 1998, according to BTW Media’s review of the case. In 2016, federal regulators and the Los Angeles City Attorney fined the bank a combined $185 million after employees opened more than 2 million bank and credit card accounts that customers may not have authorized, in order to meet sales goals. About 5,300 employees were fired, the Associated Press reported via NBC News. A later third-party review raised the count to about 3.5 million potentially unauthorized accounts, and the bank dropped retail product sales goals that year.
Audit how a target gets met as well as whether it does. A target that rises every year belongs on the risk register.
A streak turns an honest report into a loss
In March 2012, Richard Fairfax, then OSHA’s Deputy Assistant Secretary, sent a memo warning that prize drawings and team bonuses tied to injury-free periods can discourage workers from reporting injuries, especially when one report costs a whole group its reward. Such programs, he wrote, can violate the anti-retaliation provisions of the Occupational Safety and Health Act. OSHA recommended rewarding participation instead, such as hazard and near-miss reporting or work on safety committees.
Days-without-incident counters, unbroken on-call records and zero-defect streaks follow the same logic as the closed ring. Reward the reporting.
The metric sees what its sensor sees
Tickets closed, pull requests merged and deals booked show up on a dashboard. A senior engineer’s week spent mentoring two new hires does not, and neither does an on-call rotation that handled three outages or a team member whose parent is in the hospital. Like the watch, the system learns about context only when someone tells it. One-on-ones are where a manager hears it, and the plan should change once they do.
Short windows erase context
In the Adams trial, goals came down when steps came down. Most planning cycles never lower a target. Look back over a trailing quarter or half instead of a single sprint, and decide in advance which events lower the bar: an incident-heavy quarter, parental leave, a reorganization.
Expect the dip after the reward
Kivetz and his colleagues found activity fell right after customers earned their free coffee. A team that just shipped a major launch on overtime sits in the same spot. Plan a recovery period, and keep the overtime stretch out of the next baseline.
Know when the next nine costs more than it returns
Reliability engineers count availability in nines. At 99.9 percent a service can be down about eight hours and forty-five minutes a year; at 99.99 percent, about 53 minutes; at 99.999 percent, a little over five. Google’s Site Reliability Engineering book, in its chapter on embracing risk, warns that each additional increment “may cost 100x more than the previous increment,” paid for in redundant hardware and in engineers pulled off features users would see. Someone on a smartphone that works 99 percent of the time, the chapter notes, cannot tell a 99.99 percent service from a 99.999 percent one. Its worked example puts the value of moving a service with $1 million in revenue from 99.9 to 99.99 percent at $900, so any improvement costing more than that loses money.
Google’s answer is the error budget. Product and reliability teams agree on a target, a monitoring system measures actual uptime, and the gap between the two becomes the quarter’s allowance for failure. With budget left, releases ship. Once it runs low, the team slows down and works on stability. The target gets picked before the quarter starts and stays put until it ends.
A Move goal has no budget like that, and neither do most targets in an engineering org. Before funding the next nine, the next point of velocity or the next percent of quota, price it against what a user, a customer or the business would get from it.
Pair a rising rate with something people can count on
Lincoln Electric in Cleveland has paid production workers by the piece since the 1910s. In 1996, The New York Times reported that the company had never cut pay as productivity rose. By 2011, PBS NewsHour found piece rates being cut regularly to reflect productivity gains, and a pieceworker told the program that nobody wanted their price cut but there was usually a reason. Lincoln paired its rates with a long-standing policy of guaranteed employment, and the system has run for most of a century. When a target has to rise, give the reason and offer something steady in return.
Define enough, including for yourself
A stable service that is cheap to run has hit its goal, and so has a steady performer working at the level they chose. Name that level and recognize the people who keep it. The hedonic treadmill applies to managers too, so decide in writing what enough looks like this year, in scope, team size and hours, before the next title becomes the target.
A Deadline That Ends
November 19, 2026, is the date Rockstar Games settled on for Grand Theft Auto VI, after first promising the game for 2025 and then for May 26, 2026, per TechRadar. Each delay came with the same explanation, that the extra months were for polish. On August 27, Netflix premiered Grand Theft Auto VI: An Extended Look, about twenty-six minutes of gameplay, six hours before it reached YouTube; according to Inven Global, no game trailer had premiered exclusively on Netflix before. Players switch between the two leads, Jason and Lucia, mid-mission, and the footage moves through robberies, basketball, scuba diving and off-road racing across the state of Leonida. If the Extended Look is any indicator, the game is going to be remarkable.
Fixed in advance, a release date moves only when the people building the thing decide the work isn’t done. Once it arrives, it’s over. On November 19 the wait ends, and nobody proposes a bigger launch the following Monday.
That was close to what the man on Apple’s forum wanted from his watch, a number he had chosen, met and kept. He was offered a way to turn the notifications off. Next Monday the watch will send its summary anyway, and the number in it will be a little higher.
Reader discussion