Goal-gradient effect

The goal-gradient effect describes how motivation rises as people get closer to finishing something. The idea started in animal behavior research and was later shown to hold for people using loyalty cards and reward programs. In interface design it explains why progress bars, checklists and "2 of 5 steps done" labels help people complete multi-step tasks.

Where it comes from

The behaviorist Clark Hull proposed the goal-gradient hypothesis in 1932: the tendency to approach a goal increases with proximity to the goal. In a 1934 follow-up, Hull found that rats in a straight alley ran progressively faster as they moved from the start box toward the food.

Ran Kivetz, Oleg Urminsky and Yuhuang Zheng tested the idea on people in a 2006 Journal of Marketing Research paper. Their findings included:

• Members of a real café reward program bought coffee more often the closer they were to earning a free one.
• People rating songs for reward certificates visited more often, rated more songs per visit and kept going longer as they neared the reward.
• The illusion of progress worked too: customers given a 12-stamp card with 2 "bonus" stamps already filled completed the 10 required purchases faster than customers given a plain 10-stamp card.

The authors modeled effort as a function of the proportion of the original distance still remaining, which is why a head start feels meaningful even when the absolute amount of work is unchanged.

Why it matters

Many products depend on people finishing something: onboarding, a profile, a checkout, a course. Showing how far someone has come, and how little is left, gives them a reason to push through the last steps, which is often where drop-off happens.

How to apply it

Do:

• Show progress for any multi-step flow, using steps or a bar that reflects real remaining work.
• Count steps the user has already done (signing up, connecting an account) toward the total, so they start with visible momentum.
• Keep the final steps short. If the end is near, make it look and feel near.
• In an AI writing tool's setup flow, show "3 of 4 done" after the user picks a tone and uploads a sample, rather than a generic spinner.

Don't:

• Fake progress that jumps backward or stalls at 99%. Users notice, and trust drops.
• Add steps just to make the bar feel fuller.
• Hide the total number of steps. People cannot feel close to a goal they cannot see.

Common mistakes

• Treating it as a pure growth hack. Pre-filled progress works, but using it to push people into purchases or sign-ups they would not otherwise make drifts toward deceptive patterns.
• Long tails. A bar that moves quickly early and then crawls through a long final stage undoes the effect.
• No reward at the end. The effect is about approaching a goal. If finishing gives the user nothing they value, progress indicators only add noise.

In AI products

Agents and long-running AI tasks often run for minutes with no sense of how much is left. Breaking the work into named stages (searching, reading sources, drafting, checking) gives people a progress signal and makes them more willing to wait for completion. For AI onboarding, a short checklist that already counts the first prompt as done uses the same head-start effect Kivetz and colleagues found with the coffee cards. Keep the stages honest: if the agent's actual step count is unknown, show what it is doing now rather than an invented percentage.