Stretch Goals: When They Inspire and When They Backfire

Stretch goals, targets set so far above current results that nobody can yet see the route, tempt most the organizations least able to afford the risks. Sim Sitkin and four colleagues named this the paradox of stretch goals in a 2011 Academy of Management Review analysis, and their argument offers one reason a target that lifts one company can sink another.¹

Stretch goals inspire when effort matters more than know-how nobody has yet, when spare time and money can absorb a miss, and when falling short is safe to admit. Without those conditions they tend to narrow focus and invite risky bets and shortcuts. When all three hold, set the stretch above a commit goal the team expects to reach, give brand-new work a learning goal first, and pay bonuses on the commit level only; the basic goal-setting steps still apply to the commit goal itself.

What is a stretch goal?

A stretch goal is a target so far above current performance that the people asked to reach it cannot yet see how. Sitkin and colleagues describe stretch goals as seemingly impossible goals that organizations set out to reach.¹ Google’s re:Work guide, written for teams that use OKRs, places them just beyond the threshold of what seems possible and counts reaching about 70 percent of them as success.²

That separates a stretch goal from the hard goals of most goal-setting research. A hard goal asks for more of a method you already know: a sales rep who closes ten deals a month aims for fourteen. A stretch goal asks for a method nobody has found yet. A sales team told to double revenue within a year, with no new staff, has to invent a different way of selling.

In a national survey reported in 2023 by Kelly See, Chet Miller and Sitkin, roughly half of respondents across industries said their current employer had used a stretch goal in the past 12 months. Its abstract, all we could see of the paper, names neither the country nor the sample size.³

Before you call a target a stretch goal, ask whether anyone on the team can describe a route to it. If nobody can, the target is asking for invention, and the conditions below decide whether that request is fair.

Hard goals raise effort, up to a ceiling

Edwin Locke and Gary Latham built goal-setting theory from decades of experiments, and it holds that specific, difficult goals raise effort and performance more than easy goals or a plea to do your best. Their 2002 review adds the ceiling: performance levels off or falls once a goal outruns people’s ability or their commitment to it lapses.⁴

Commitment counts for most on hard goals, which demand more effort and offer lower odds of success, and people need to see their progress in order to adjust. The review traces commitment to two sources: a sense that the goal is important, and confidence that it can be reached, which psychologists call self-efficacyself-efficacy: A person's belief in their ability to carry out a particular task or reach a particular goal. It is specific to the task, so someone can feel capable at work and unsure with money, and it goes with how readily people start and keep going.Full entry in the glossary.⁴

A stretch goal presses on each of those conditions at once. It sits past current ability, it lowers the odds that commitment rests on, and early feedback mostly reports how far behind you are. A warehouse team told to halve picking errors this quarter may see the count barely move by week three, and the dashboard meant to motivate starts to read as a verdict. The same tension runs through the debate over the “realistic” in SMART goals.

Myth

The higher the target, the harder people will work.

Fact

Effort rises with difficulty only while people have the ability and still believe the goal is reachable; past that point, commitment and performance drop.

Treat ability and commitment as the two gauges to watch. Once people stop believing a target is reachable, it no longer raises effort, however inspiring it sounded at the kickoff.

Are stretch goals good or bad? It depends on recent results and slack

Stretch goals are neither good nor bad in general. In a 2023 paper, See, Miller and Sitkin conclude that they can be valuable only when an organization has both strong recent performance and slack resources, meaning spare money, time and people to put into the attempt; under other conditions, they write, stretch goals tend to be demotivating and destructive.³

In a 2017 article for Harvard Business Review, Sitkin, Miller and See observe that organizations with recent wins and slack seldom use stretch goals, while businesses in trouble often adopt them in a desperate attempt to turn around, and that those attempts nearly always fail. Their example is Yahoo: after Marissa Mayer took over the struggling company in 2012 she announced several very ambitious targets, among them double-digit annual growth, and by the summer of 2016 had fallen far short on most of them, double-digit growth included.⁵

Recent winsWeak recent resultsSlack resourcesNo slack

Slack resources, Recent wins: Stretch can pay offroom to try and fail

Slack resources, Weak recent results: Riskyresources but no momentum

No slack, Recent wins: Riskymomentum but no cushion

No slack, Weak recent results: Most tempted, most at riskwhere stretch goals tend to be adopted

Gary and colleagues, summarizing earlier research in 2017, describe two ways the paradox plays out. Falling short of a goal pushes people toward new, untested strategies as long as they are safely above the point of collapse, so a far-off target raises risk-taking. And while some people stay committed, others lose confidence after repeated failure, and with it commitment and learning.⁶ Our reading is that slack keeps an organization further from that point of collapse, so its experiments can fail without ending it.

Before you set a stretch goal, answer two questions honestly: have we won recently, and what could we lose and still carry on? A team that just shipped on time with budget left over sits in the top-left cell. A team behind plan and short of people sits where stretch goals are most tempting and most dangerous. The same two questions apply to one person’s quarter, and personal OKRs written for a season with no slack are better shrunk or postponed.

What happens when people chase a stretch goal

In two experiments published in 2017 by Michael Shayne Gary, Miles Yang and Philip Yetton of UNSW Sydney with John Sterman of MIT, stretch goals did not raise typical performance in a business simulation. They spread results out instead.⁶

The study

Limited evidence

Ten simulated years running an airline, under a stretch or a moderate profit goal (2017)

Teams of managers on an Executive MBA course ran a simulated airline over ten simulated years, randomly assigned either a moderate profit goal or a stretch goal set at the level the top tenth had reached in pilot tests. About one in five stretch teams met their goal, against about three in five moderate teams, and more stretch teams went bankrupt (43 versus 33 percent). Median profit was no higher under stretch goals, and results varied far more. A second experiment with 59 students found the same pattern, and their commitment to the stretch goal fell from round to round.⁶

The stretch targets were reachable, and well below the best possible result, yet most participants either chose growth plans that ended in bankruptcy or swapped the target for a lower goal of their own or for plain survival.⁶ The study was small and artificial: a few dozen teams playing a gamelaboratory experiment: An experiment run under conditions the researchers set up and control, often with students doing short tasks. Control makes cause and effect easier to pin down, but the tasks are simpler than real life, so results do not always carry over to work or home.Full entry in the glossary scored only on profit, with no careers, customers or colleagues at stake.

1. Moderate goal: results bunch together, and many land near or past the goal
2. Stretch goal: results spread out: a few go far ahead, many fall short or give up the target, and the typical result does not improve

Ask what the bottom of that spread would mean for you. If a few big wins can pay for many misses, as in a portfolio of small product bets, the spread may be worth it; if one bad year ends the team, it is not.

Narrow focus, shortcuts and lost learning: how stretch goals go wrong

Very hard goals can narrow attention, push people toward risky bets and dishonest shortcuts, and crowd out learning, Lisa Ordóñez and three fellow management researchers argued in “Goals Gone Wild”, a 2009 critique.⁷ Three of its points bear directly on stretch goals.

The first is risk. In 1976 the chairman of Continental Illinois, a US bank, announced that within five years its lending would match any other bank’s. To get there it moved from conservative corporate lending to an aggressive pursuit of borrowers and bought loans from smaller banks that had made very risky ones; when those loans defaulted, the government had to bail it out.⁷

The second is learning. On a complex task, the authors argue, a specific, challenging goal can keep people from trying the alternative methods that would teach them the task, and managers may struggle to judge when a task is complex enough to call for a learning goal instead. The third is cheating: they point to a study in which people given a specific, challenging goal misrepresented their performance more often, most of all when they had narrowly missed it.⁷

Locke and Latham answered in the same year, saying the critique leaned on anecdotes in place of research.⁸ Company cases, including the one below, cannot prove that goals cause misconduct.

A widely reported recent case comes from US banking. In 2016 the Consumer Financial Protection Bureau (CFPB), the US consumer finance regulator, said that Wells Fargo employees, spurred by sales targets and compensation incentives, had opened accounts customers may not have authorized; the bank’s own analysis found more than two million such deposit and credit card accounts.⁹ The bank fired about 5,300 employees over fraudulent sales between 2011 and 2016.¹⁰ In 2020 it agreed to a combined US$3 billion settlement with the US Justice Department and the Securities and Exchange Commission.¹¹

A 2013 Los Angeles Times investigation had already described sales quotas so aggressive that some were mathematically impossible.¹⁰ The regulator warned that financial incentive programs carry serious risks if not monitored carefully.⁹ Ordóñez and colleagues name lax oversight, financial incentives for meeting targets and a weak ethical culture as the conditions that turn goals into cheating.⁷ When you set a stretch target, decide at the same time who will look at how results were reached, and look hardest in the final weeks, when people are just short.

When stretch goals are worth setting

Stretch goals are worth setting when three conditions hold: effort and search matter more than a skill nobody has yet, there is slack to absorb a miss, and failure is safe to report. A goal that fails any one of them works better set lower.

Skill comes first. Locke and Latham’s review explains why: while people are still working out how to do a task, pressure to hit a number can make them hunt for methods in a scattered way and miss what works, but once they know the methods, a high target makes them more likely to apply them.⁴ A call center that has already settled on a reliable way of handling complaints can chase a big jump in calls resolved; a team moving to new software in the same quarter is better served by a goal to learn the tool.

Slack is the second condition, as the paradox research argues.³ Safety is the third. In the 2009 critique, General Electric’s Steve Kerr is quoted saying most organizations do not know how to manage stretch goals; he advises avoiding goals that raise stress, not punishing failure and giving people the tools to reach ambitious goals.⁷ Google’s guide asks leaders to state the threshold for success plainly, so that missing part of a stretch reads as normal.²

How do you set a stretch goal at work?

Set a stretch goal at work by checking the conditions first, then surrounding the target with a realistic commitment, a learning goal for new work, pay rules that do not hang on the stretch, and a review of how results were reached. Nobody has tested these five steps together; each rests on its own source.

1. Check recent results and slack before writing the number

Look at recent periods and at what spare budget, time and people exist. The 2023 paper finds stretch goals valuable only when both are in place.³ If either is thin, our reading is that a hard but reachable goal is the safer choice.

2. Pair the stretch with a commit goal you expect to hit

In the 2017 simulations, many people facing an out-of-reach target quietly replaced it with a lower goal of their own.⁶ Set that lower goal openly: a commit level the team expects to reach and will be judged on, with the stretch above it as the aim. Google’s guide makes the same point about saying what counts as success.²

3. Give unfamiliar work a learning goal first

When the route is unknown, make the first goal about finding it. Locke and Latham cite an experiment on a complex task in which people given a specific, difficult learning goal, such as finding a set number of strategies, did better than people simply asked to try their best.⁴ Switch to the performance number once a method works.

4. Pay bonuses on the commit goal

Locke and Latham’s review reports that an all-or-nothing bonus tied to a very hard goal can drag performance down: as the payout starts to look unreachable, people lower their own goal and their confidence, and output falls with them. With moderately hard goals, or with pay for each unit produced, the drop did not appear.⁴

5. Review how the numbers were reached

At each check-in, ask how results were produced as well as whether they arrived, and treat a miss as information. Keep the near-miss weeks in view, where the experiment cited in the critique found the most misreporting.⁷

Before you set a stretch goal

• We have had recent wins, and there is spare budget, time or people
• A commit goal sits below the stretch, and everyone knows which one counts
• New work starts with a learning goal
• No bonus depends on hitting the stretch
• Someone reviews how results are reached, not only whether

The bottom line

A stretch goal is a bet that pays off mainly for teams that are already winning and have room to absorb a miss. Before you set one, look at the last few quarters and the spare budget, write a commit goal beneath it, and keep pay tied to the commit level. Then name the person who will check how the numbers are reached.