A member with nine stamps on a ten-stamp card comes back sooner than a member with three. That pull toward the finish line has a name, the goal gradient effect, and researchers have measured it in rats, coffee shops and music websites. For a loyalty team it explains why tier meters and punch cards work. It also explains what goes wrong right after the reward.

What is the goal gradient effect?

The goal gradient effect is the tendency to put in more effort as a reward gets closer. The behaviorist Clark Hull proposed the idea in 1932. He predicted that animals in a maze would move at a quicker pace as they approached the goal. In 1934 his rats ran a straight runway and ran faster the nearer they got to the food (Kivetz, Urminsky and Zheng, 2006).

For people, the idea stayed thinly studied for decades. In 2006 Ran Kivetz, Oleg Urminsky and Yuhuang Zheng tested it in the Journal of Marketing Research. They used a real campus café and a music-rating website. Their results made the effect usable for anyone who runs a reward program.

Their model is simple. Effort depends on the share of the original distance still left to the goal (Kivetz et al.). By that logic, two stamps left on a ten-stamp card and 100 Stars left on a 500-Star climb sit at the same distance, 20%. Members respond to how close they are, and the raw count matters less.

What did the coffee card study show?

Customers bought coffee faster as they neared the free one. The café gave out ten-stamp cards, one stamp per visit, and the team tracked 949 completed cards covering about 10,000 purchases. The time between purchases fell by 20%, or 0.7 days, from the first stamp to the last. Cards took 24.6 days to finish. At the speed of the first purchases, they would have taken 29.4 days (Kivetz et al.).

The team added a control. They enrolled 42 customers in a fake program whose cards earned nothing. Those customers did not speed up, and their pattern ran the other way (same paper). The speed-up came from the reward and not from habit or the calendar.

StudySettingWhat it found
Hull (1934), as cited by KivetzRats in a straight runwayRats ran faster the closer they were to the food
Kivetz, Urminsky and Zheng (2006)Campus café, ten-stamp cardPurchase gaps shrank 20% toward the free coffee; no speed-up without a reward
Kivetz, Urminsky and Zheng (2006)Music-rating website with reward certificatesPeople visited more often, rated more songs per visit and stayed longer as they neared the reward
Bonezzi, Brendl and De Angelis (2011)Experiments on progress monitoringMotivation can be higher far from and close to the goal and lower halfway
Koo and Fishbach (2012)Experiments on progress framingEarly on, stress what is done; near the end, stress what is left
Cheema and Bagchi (2011)Experiments on visual progressProgress bars help only when people are close to the goal, and less when the goal is split into subgoals

Is the effect always a straight climb to the goal?

No. Andrea Bonezzi, Miguel Brendl and Matteo De Angelis found a dip. In 2011 they showed that motivation can be higher when people are far from the goal or close to it, and lower around the halfway mark. Whether people track distance from the start or distance to the end changes the shape of the curve (Bonezzi, Brendl and De Angelis, 2011).

Minjung Koo and Ayelet Fishbach gave a practical rule in 2012. At the start of a goal, telling people what they have done, such as 20% completed, beats telling them what remains. Near the end, 20% remaining works better than 80% completed. People respond to whichever number is smaller (Koo and Fishbach, 2012).

Display matters too. Amar Cheema and Rajesh Bagchi found that progress bars lift effort as people approach a goal. The lift appears only when people are close, and it shrinks when the goal is cut into subgoals (Cheema and Bagchi, 2011).

What happens after the reward?

Speed drops, then returns. Kivetz and colleagues followed 110 café members who finished one card and began a second. Their first two gaps on the new card were 3.1 and 2.7 days. Their last two on the old card were 2.2 and 2.1 days. The new gaps matched the first two on the first card, 3.2 and 2.8 days (Kivetz et al.). Members slowed after the reward and sped up again as the next one came into view.

The same paper links the effect to retention. People who accelerated more strongly toward their first reward were more likely to earn a second and came back sooner. So a program should have the next goal ready the moment the last one pays out.

How do loyalty programs use the goal gradient effect?

Anywhere a program sets a threshold and shows the member where they stand. Starbucks sets a member’s level by Stars earned over 12 months. A member reaches Gold at 500 Stars, then Reserve at 2,500 new Stars earned after reaching Gold. Members can track their Level Star Balance in the app (Starbucks Rewards Terms).

Delta uses dollars. For 2027 status, Delta lists thresholds of $5,000 for Silver, $10,000 for Gold, $15,000 for Platinum and $28,000 for Diamond in Medallion Qualification Dollars. The count resets each calendar year (Delta Air Lines). A year is long enough that members spend months far from the line and then a few weeks close to it. The pull is strongest in those last weeks.

The research gives three design rules. Show the meter. Show it in the frame that suits the distance, done early and left late. Split a very long journey into steps so there is always a near goal.

How can a loyalty team test it?

Log three things for each purchase: the member’s distance to the threshold, the date and the spend. Plot the days between purchases against distance. A real gradient slopes down as distance falls, and a flat line says the meter changes nothing. Then run a holdout. For illustration, a program has 40,000 members in a 500-Star cycle, and 6,000 sit within 75 Stars of the next tier. The team shows a progress meter to 3,000 of them and holds it back from 3,000. In the next 30 days the meter group makes 3.0 visits per member against 2.6 for the holdout. That is 0.4 visits more, or 1,200 visits across the group.

Then check the 30 days after. If the meter group makes 0.3 fewer visits than the holdout in that window, the net gain falls to 0.1 visits per member, or 300 visits. The Kivetz slowdown after the reward is why the second window matters. Count the net, and judge the meter on that.

How GRAVTY handles the goal gradient effect

Tier lines and meters are program rules. GRAVTY programs run badges, milestones and challenges. SHARE at Majid Al Futtaim is one. Magenta Moments at Deutsche Telekom is another. Patented Visual Rules score each behavior as it arrives. The reward lands in the member account. Teams change these rules without code. See the gamification guide and the technology page.

Frequently asked questions

Who discovered the goal gradient effect?

The behaviorist Clark Hull proposed the goal gradient hypothesis in 1932. In 1934 his rats in a straight runway ran faster the closer they got to the food (Kivetz, Urminsky and Zheng, 2006). Ran Kivetz and colleagues carried the idea to human customers in 2006.

What is an example of the goal gradient effect?

A coffee loyalty card is the classic example. In a campus café, the time between purchases fell by 20% as customers moved from the first stamp to the tenth (Kivetz et al.). Members bought faster the closer they got to the free coffee.

What is the difference between the goal gradient effect and the endowed progress effect?

The goal gradient effect says effort rises as the goal nears. The endowed progress effect, named by Joseph Nunes and Xavier Drèze in 2006, says people given artificial advancement toward a goal show more persistence (Nunes and Drèze, 2006). Kivetz and colleagues tested a close cousin with a 12-stamp card that came with two free stamps. Its owners finished in 12.7 days, against 15.6 days for a plain ten-stamp card (Kivetz et al.).

Does the goal gradient effect work with progress bars?

Yes, when people are already close to the goal. Cheema and Bagchi found that visual progress lifts effort as people near a goal. The lift fades when the goal is split into subgoals (Cheema and Bagchi, 2011). A bar at the start of a long goal does less.

Can the goal gradient effect backfire?

Yes, because effort falls after the reward. In the café, the first gaps on a second card were as slow as the first gaps on the first card (Kivetz et al.). A program that sets one big goal and then goes quiet loses the speed it gained.