Most people have never read the terms and conditions of their credit card(s). Research suggests that, even among those who do access such documents, the time spent falls far short of what would be needed to understand the agreement. One study of more than 130,000 visits to 90 online software companies found that only 0.05% of visitors accessed the terms and conditions pages at all, and among those who did, fewer than 7% of the words in those contracts would have been read based on the time spent on those pages.

That gap between what financial institutions disclose and what customers absorb sits at the heart of new research examining whether a single text message, sent immediately after a credit card transaction, could change financial behaviour. The answer is: yes, and the implications extend well beyond credit cards.

What happens when no one reads the fine print

When an Australian bank customer used their credit card for a cash advance, a gambling transaction, or a money order, two costs applied automatically. The first is a fee of $4 or 3% of the transaction amount (up to a maximum of $300, whichever is greater). The second is a premium interest rate of 21.24% per annum, accruing from the transaction date, with no interest-free period. This rate is approximately 8 percentage points higher than the rate for ordinary purchases.

These so-called high-cost transactions cover five categories: betting and gambling, manual cash disbursements, automated cash disbursements, money orders, and non-financial money orders. Yet many customers who make these transactions were likely unaware of the fees and elevated interest they had just triggered, even though that information was available in their credit card contracts.

The researchers behind Nudging a Second After, published in the Journal of Banking and Finance, asked whether informing customers about these costs immediately after each transaction, rather than beforehand, would prompt them to adjust their behaviour.

Traditional pre-transaction warnings appear insufficient, as they are often overlooked or not fully processed at the moment of decision-making. This prompted both the research co-authors and the bank that provided the data to explore whether providing real-time feedback immediately after a transaction could improve customer awareness and influence subsequent behaviour.

The aim of the study was to determine whether “ex-post” nudges can reduce costly financial actions and encourage more responsible repayment patterns. From the bank’s perspective, this approach also reflects a concern for customer welfare and an effort to reduce harmful borrowing patterns, offering a potentially scalable, low-cost tool to improve consumer financial outcomes.

The experiment: a text message as a behavioural tool

The research paper was co-authored by Professor Kadir Atalay from the School of Economics at The University of Sydney, Dr Hanlin Lou, a research fellow at the UNSW Centre for Population Ageing Research (CEPAR), and Professor Robert Slonim from the Economics Discipline Group at the University of Technology (UTS) Sydney.

Subscribe to BusinessThink for the latest research, analysis and insights from UNSW Business School

The study examined a field experiment conducted by a major Australian bank from October 2017 to October 2018, using transaction-level data from 325,195 credit card accounts across the treatment and control groups, with one year of pre-experiment data used to confirm that the two groups were statistically equivalent before the trial began. Customers in the treatment group received an automated text message (in-app notification) immediately after each high-cost transaction; customers in the control group received no such notification.

What the data revealed

Customers who received notifications reduced their subsequent high-cost transactions by approximately 6% compared to those who did not. Among customers who had made at least one such transaction during the experiment (and were therefore directly exposed to the nudge), the reduction was 5.7%, equivalent to roughly 0.4 fewer high-cost transactions per customer per year.

The effect was most pronounced for betting and gambling transactions, which fell by 14.5% among those who received notifications. Cash advance transactions fell by 9.7%. The contrast between the two categories was telling. Cash advance users are more likely to be using their credit cards because they have no other source of funds (a situation described by economists as a liquidity constraint).

For those customers, a cost notification may not have changed behaviour, because the transaction was driven by necessity rather than unawareness. Gambling transactions, by contrast, are typically made online using a registered credit card regardless of a customer's cash position. For that group, a lack of awareness of the high-cost nature of the transaction was a more plausible explanation for why the nudge proved effective.

Notifications also changed repayment behaviour. Customers who received the text were 5% more likely to make a credit card repayment on the same day as the transaction, an effect driven substantially by those who had made gambling transactions. The treatment group was also 1.2% more likely to have fully paid off the high-cost transaction amount on the same day it was made, rising to 1.5% by the third day.

Why timing matters in behavioural nudges

The conventional wisdom on nudging people toward better decisions is to provide information just before they act: reminders about loan repayment deadlines, disclosures about overdraft fees before a customer goes into the red, or warnings about the cost of credit before a purchase is made. This approach assumes that well-timed information can prevent a costly decision.

"Real-time push notifications can serve as an effective and scalable behavioural intervention"

HANLIN LOU

But this logic only works when the decision can be predicted. Many financial behaviours that carry penalties cannot be forecast with precision. Sending a general warning to a broad customer base about potential future costs risks being ignored, forgotten, or (in a twist the paper explores in some depth) inadvertently drawing attention to options that customers had not previously considered.

The researchers argued that providing information immediately after a costly transaction leverages a different set of psychological mechanisms. The customer has just experienced a financial loss, making it more salient (i.e., more front-of-mind and attention-grabbing) than an abstract warning about a hypothetical future event.

There is also the effect of what behavioural economists call loss aversion: the tendency for people to respond more to losses than to equivalent gains. A notification confirming that money has just been taken, and that a higher interest rate is now running, operates on both of these impulses simultaneously.

The reduction in high-cost transactions persisted across the full year of the experiment. The researchers also found no evidence that customers simply shifted their gambling or cash transactions to their debit cards; the behaviour changed rather than migrated.

What this means for financial services and regulation

The annual cost savings from the intervention were estimated at approximately $4.35 per affected customer, equivalent to around 1% of median monthly credit card expenditure. Across the more than 500,000 credit card accounts in Australia, estimated to make high-cost transactions each year, the total annual savings to customers exceeded $2 million. The savings were highest among younger customers, those with lower socioeconomic status, and those with a history of frequent high-cost transactions: groups that tend to be more financially vulnerable.

For financial services professionals, the research demonstrates that real-time notification systems (now a feature of most banking apps) can be deployed not just for security alerts or balance updates, but as an instrument for improving customer financial outcomes. The experiment found no statistically significant effect on account closures or on normal credit card spending, which the researchers took as an indication that the notifications did not damage the customer relationship.

The researchers noted that the notifications could be made more effective. A link to a repayment page embedded in the text message would allow customers to act immediately, rather than needing to navigate separately to their banking app. There is also scope to tailor notifications by transaction type; the research found that a notification about one category of high-cost transaction had a limited effect on behaviour in other categories.

At a regulatory level, the paper raises the possibility that mandatory real-time notification requirements could serve as an alternative to price regulation or outright bans on certain transaction fees: a policy tool that preserves customer choice while reducing the information gap that allows hidden costs to persist.

“The findings suggest that the timing of information disclosure plays a critical role in shaping financial behaviour,” said CEPAR’s Dr Lou. “Providing immediate, post-transaction feedback increases customer awareness of costs at a moment when the decision is still salient, leading to measurable improvements in subsequent actions. Real-time push notifications can serve as an effective and scalable behavioural intervention.”

As such, he concluded that financial institutions should incorporate ex-post nudges into their digital channels to promote more informed decision-making and help reduce harmful borrowing patterns.