Nearly one in three Europeans can't cover a surprise bill
In 2025, 29.2% of people in the EU lived in a household that could not face an unexpected financial expense. In Latvia the figure was 43.7%, in Lithuania 41.1%, in Germany 31.9% and in Poland 22.5%.28
Most people already know they should save. The hard part is doing it month after month, when the reward is years away and the shop is one tap away. That gap between knowing and doing is exactly where game design claims to help. So it is worth asking what the evidence actually says.
What dopamine actually does
A prediction signal, not a pleasure chemical
The most quoted line about gamification is that it gives you "a dopamine hit". The research behind dopamine says something more precise and more useful. In the classic experiments by Wolfram Schultz and colleagues, dopamine neurons fired when a reward arrived unexpectedly. Once a cue reliably predicted the reward, they fired at the cue instead and stopped responding to the reward itself. When an expected reward failed to arrive, their activity dipped at the exact moment it was due.1
In other words, dopamine tracks the difference between what you expected and what you got. It is a learning signal. Kent Berridge's work separates "wanting", which relies heavily on dopamine, from "liking", the actual pleasure, which does not depend on it.4 That distinction matters for money apps: a mechanic can make you want to open an app without making your finances any better.
Anticipation is its own signal
Dopamine neurons also show a slower, ramping response while a reward is pending, strongest when the outcome is most uncertain.2 In brain imaging, simply anticipating a larger monetary reward increased activity in the nucleus accumbens.3 Uncertainty is what makes slot machines compelling. A finance app that borrows that mechanic is borrowing from the wrong place.
What this means for design: Progress signals work best when they carry real, new information about your situation. Random rewards make people want to come back; they do not make them better off.
The psychology of getting closer
Effort rises as the goal gets near
In a study of a real café loyalty card ("buy ten coffees, get one free"), the time between purchases shrank by about 20% as customers approached the free coffee. Customers given a 12-stamp card with 2 stamps already filled finished it in 12.7 days on average, against 15.6 days for a plain 10-stamp card. After the reward, they slowed down again.5 Psychologists call this the goal-gradient effect.
A single savings goal of €10,000 has one finish line, far away. Split into many smaller steps, it has a finish line that is always close.
A head start, if it is honest
The same principle appears in the "endowed progress effect": people who are told they are already part of the way through a task are more likely to finish it, and finish faster.6 For money, the honest version is counting progress that is already real, such as debt you have already paid down.
Specific goals beat "do your best"
Decades of goal-setting research found that specific, difficult goals lead to higher performance than vague ones, with effect sizes between d = 0.42 and 0.80.8 "Save more" is a wish. "€6,000 in an emergency fund" is a goal.
Seeing your progress works
A meta-analysis of 138 randomised studies with nearly 20,000 participants found that prompting people to monitor their progress improved goal attainment (d = 0.40), and the effect was larger when progress was recorded or reported, not just thought about.9 In a diary study of 238 knowledge workers, progress in meaningful work happened on 76% of their best days.7
Does gamification work? Small effects, honest caveats
The broad answer from the research is: yes, a little, depending on the design and the person. An early review of 24 empirical studies found mostly positive effects, but warned they were "greatly dependent on the context" and on the users.10 A later meta-analysis of gamified learning measured effects the authors themselves call small:
| Outcome | Effect (g) | Studies | Participants |
|---|---|---|---|
| Cognitive (what people learned) | 0.49 | 19 | 1,686 |
| Motivational | 0.36 | 16 | 2,246 |
| Behavioural | 0.25 | 9 | 951 |
Source: Sailer & Homner, 2020.11 In a controlled experiment, points, levels and leaderboards raised how much people did, but did not change their intrinsic motivation.13 And engagement often spikes at launch and drops a few weeks later, the so-called novelty effect.14
The one saving experiment
We found only one randomised study that tests gamified saving directly. 331 people in the UK used a savings tool for four weeks. Those with a progress bar, badges and a leaderboard saved 27.3% of their weekly goal, against 22.9% for a plain version, and reached 108.9% of their goal by the end, against 91.4%.12 The effect is real but modest (d = 0.22), the study was short, and savings were self-reported. Nobody has yet run a long-term trial of a gamified budgeting app. Anyone who promises a specific percentage of extra savings is going beyond the evidence.
What already works for saving
Commitment
In the Philippines, bank clients were offered a savings account that locked their money until a goal they chose themselves. After twelve months, the group offered it had average balances 81 percentage points higher than a control group, even though only 28% of those offered took it up.15 In the US, the "Save More Tomorrow" plan let employees commit in advance to saving part of future pay rises. 78% of those offered it joined, and their saving rate rose from 3.5% to 13.6% over 40 months.16
Reminders that name your goal
In field experiments in three countries, monthly reminders raised the amount saved by 6%. Reminders that mentioned the person's own savings goal raised it by about 16%. Generic reminders did nothing.17
Prize-linked savings, and why we don't use chance
Savings accounts that enter depositors into prize draws increased total savings by 1% of annual income in South Africa, largely by replacing lottery spending.18 In Mexico they led to 41% more accounts being opened, still in use five years later.19 They work. But they work through uncertainty, the same signal that makes gambling compelling.2
Streaks that forgive a missed day
Habits form slowly and at very different speeds: in one study, the time to reach near-maximum automaticity ranged from 18 to 254 days, and missing a single day did not materially affect the process.20 Streaks themselves change behaviour: across seven studies, people shown an intact streak engaged more afterwards than people shown a broken one, even when their actual past behaviour was identical. The drop was smaller when people could "repair" the streak.21 Duolingo has reported that allowing streak freezes increased daily active learners,22 though that is company data, not an independent study.
Where gamification backfires
Rewards can crowd out motivation
A meta-analysis of 128 experiments found that tangible rewards given for doing a task reduced people's freely chosen interest in it afterwards (d = −0.28 to −0.40). Positive feedback, on the other hand, increased it (d = +0.33).23 The difference is whether a signal feels like a bribe or like information about how you're doing.
A third of feedback makes things worse
Across 607 effect sizes, feedback improved performance on average, but more than a third of feedback interventions reduced it, especially when feedback pulled attention towards the self rather than the task.24 "You did badly this month" is about you. "€120 went to your goal this month" is about the task.
Losses loom larger
People weigh losses more heavily than equal gains.25 A system that takes levels away after an expensive month would likely hurt more than the original level-up felt good, and would punish the moments when people most need encouragement.
Money stress and the ostrich effect
Investors log into their accounts 9.5% less after market declines: when the news is bad, people stop looking.27 Research on whether financial pressure itself reduces mental bandwidth is mixed. One well-known study found farmers scored worse on cognitive tests before harvest than after, while a US study found no difference around payday.26 Either way, an app that adds blame to money worries gives people one more reason to look away.
How the evidence shaped Groworth's rules
Groworth turns your own goal, whether an emergency fund, a home deposit or being debt-free, into 100 levels. These rules come straight from the research above:
Your own, specific goal. Levels measure progress towards the number you chose, never a generic score.89
XP only for decisions. Saving, investing and paying off debt earn XP. Income arriving earns none, and neither do market price moves, in either direction: being paid is not a decision, and a market crash shouldn't erase your progress.2327
Levels never go down. An expensive month shows up in your spending limits, not as a lost level.2425
No leaderboards and no chance. Your progress is compared only with your own goal, and nothing is random.224
Reminders only when there is something to do. At most one a day, and only when new transactions are waiting to be sorted.17
The tools behind the levels are free: a debt payoff calculator, an emergency fund calculator and an honest comparison with 31 other money apps.
What we don't know yet
The evidence for each ingredient, goals, visible progress, commitment and reminders, is strong. The evidence that putting them together in an app makes people save more over years is not there yet: the one direct experiment lasted four weeks. Novelty fades, and small effects are still small. Gamification is not a substitute for a plan; at best it keeps a good plan in front of you long enough for it to become a habit.