Goal Setting and Saving in the FinTech Era

GAICIL Insights

The cheapest commitment device ever invented may be a box in an app that says "goal." Goal Setting and Saving in the FinTech Era, by Antonio Gargano and Alberto G.¹ Rossi, studies a classic gap between intention and follow-through. People intend to save, and many fall short. The question is whether the ability to set saving goals in a FinTech app closes that gap. With a credible causal design, the answer is yes.

Identification exploits random assignment of users into a group of beta testers who could set goals and a group who could not. Using difference-in-differences, the authors find that setting goals raises individuals' savings rate. Crucially, increased savings within the app do not come at the expense of reduced savings outside it. Goal setting also helps the individuals previously identified as having the lowest propensity to save. The skeptic's two favorite objections, money shuffling and help only for the already-thrifty, both fail.

The channels come from matching survey responses to behavior. Monitoring matters, consistent with models in which people dislike falling short of a goal and in which goal setting helps them follow through by increasing attention. Concreteness plays a role too. Saving goals are soft, self-designed commitment devices: they impose no penalties, yet they change behavior. No lock on the account, no fee for failure, just a target and a progress bar.

This captures a central promise of artificial intelligence and digital tools in personal finance. Apps make goals salient, track progress continuously and remind users when they drift, attacking exactly the attention and monitoring problems the paper identifies. A tool that costs little to deliver can raise savings for millions, including those who struggle most. The marginal cost of the intervention is close to zero; the marginal benefit is a fatter emergency fund.

The no-shuffling result deserves emphasis. A skeptic might suspect users merely moved money between accounts. The evidence rules that out, indicating genuine increases in saving and strengthening the case for digital commitment devices as tools for household financial resilience.

Reaching low-propensity savers is the rarer achievement. Many interventions mostly help those already inclined to save. Goal setting reached those least likely to save, suggesting simple, well-designed tools can narrow gaps in financial behavior. AI could go further: personalizing goals, adjusting reminders to individual patterns and offering encouragement at the moments users are most likely to lapse. A reminder sent on payday may beat a reminder sent at random.

Softness is a feature. Hard commitments, such as accounts restricting withdrawals, can work and can also deter participation. Self-designed goals preserve freedom while harnessing psychology. This fits a philosophy of empowering individuals with information and tools, leaving choices in their hands. Paternalism is optional; attention is the active ingredient.

The study also shows FinTech platforms as evidence engines. Randomized rollouts of new features, common in technology firms, create natural experiments. Collaborations between researchers and platforms produce rigorous findings about what works, informing product design and public policy. As AI features proliferate, such experiments will be essential to separate effective innovations from ineffective ones. Every feature launch is a potential trial.

Policy follows. Encouraging innovation in consumer financial tools can serve public goals. Raising household saving is a long-standing objective. Digital tools that help people set and monitor goals contribute at low cost, without mandates. Regulators can support them by ensuring data protection and transparency while allowing experimentation.

At bottom, small design choices in digital tools can produce meaningful behavior change. Setting a goal, monitoring progress and paying attention helped users save more, including those who had saved least. As AI makes financial tools more personalized and responsive, such gains can grow. Technology can help people do what they already intend to do.

Before designing another tax incentive for saving, test a goal feature. The evidence here suggests a progress bar can move behavior that decades of exhortation left unchanged, for a fraction of the cost.