With the two biggest continental football tournaments now over, we decided to look at the numbers once again to see if we could answer the question we posed six weeks ago: Would fraudsters win big at the UEFA Euro 2024?
One of the obvious conclusions from previous tournaments is the reduction in digital banking activity in general during matches. We saw this trend continue into this year’s tournament, particularly for the matches during the latter stages:
| Stage | Spain | UK |
|
Semi-finals |
20% | 19% |
|
Final |
32% | 26% |
There’s no doubting there’s a direct relationship between digital banking activity and football matches, and as matches become more significant, more people watch on TV, leading to an even greater reduction in digital banking activity.
As we explore further, we see changes in behavior across the general population.
Device usage
During the group stages, we saw users change their device preference. It goes without saying that mobile devices are the preference for the vast majority. For example, in the UK, 93% of all connections for digital banking sessions take place from mobile apps, with a further 1% connecting from web browsers on a mobile device. In Croatia, we see more than 98% of activity originate from mobile devices, and in Spain this is around 80%.
However, when the game is on, we see changes in user preferences. In Spain, we saw 10% less activity on desktop devices during group-stage matches. In Croatia, desktop activity declined by 12%, with peaks of 19% during the Croatian side’s final group match. Meanwhile, in the UK, the behaviour was different, with a higher preference for desktop devices, with increases between 5 and 10% during Three Lions group matches (although for the final, we saw the same trend we observed in other European countries, with a 15% shift to mobile).
Betting influence
A 2023 survey by YouGov showed that football is the No. 1 sport for online sports gamblers, with seven out of 10 sports gamblers in the UK having placed a bet on a football match.[1] Therefore, it’s no surprise that we see a link between significant football matches and betting, even within digital banking.
During the group stage matches, we were able to identify an 11% increase in digital banking traffic from mobile devices where betting apps were installed in the two hours prior to England and Scotland matches. During the matches, activity remained in line with previous weeks. Then, once the matches were over, there were spikes once again from such devices.
Analysing the data available for the day of the final, we see a similar increase prior to the match (9%). During the match, activity drops by 4% from devices with betting apps installed. Then, after the match is over, we observe a 27% spike.
Understanding human behavior, we believe that:
• There is an increase in activity before the match, as fans and bettors are placing their bets. With SCA, many banks push their customers through their mobile app to authorise card payments. Therefore, any increase in betting would logically lead to an increase in banking sessions from devices with betting apps.
• During the match, fans and bettors are focused on the football. There may be some population of people betting in real-time, but in general, this isn’t the norm for most watching the games.
• Once the match is over, bets are closed. Those lucky enough to have made money will be looking to cash out their balances from their betting wallets, transferring funds over to their bank accounts. Users are likely checking their bank balance to ensure funds are received.
Active calls
For the 2020 Euros (which took place in 2021), we identified a reduction in voice scams on match days. One distinctive feature of voice scams is that there is typically an active phone call while the mobile banking session takes place. Therefore, we decided to investigate whether there are any changes in behaviour in this area in 2024.
Looking at data from BioCatch’s UK customers on July 14 (the day of the final), we see a 33% reduction in the ratio of sessions with an active call during the match (compared tothe same period from the previous Sunday). This doesn’t come as a surprise, as most of the nation was watching the football match (UK viewership is reported to be at 90% share[2]).
[1] https://business.yougov.com/content/46255-which-sports-attract-the-biggest-betting-spends-in-britain
[2] https://www.sportspromedia.com/news/euro-2024-final-tv-viewership-bbc-itv-rtve/
Interestingly though, once the match was over, the ratio of active calls spiked, with a 57% increase compared with the week before. A similar trend was picked up on the 10th July (the day of the semi-final), where active call rates shot up by 99%. The emotion of the match and friends commenting on it, while checking bank accounts for cashed-out bets, is the most likely explanation of this.
In summary, we see how genuine users’ behaviour changes around these events, ranging from not using digital banking at all when matches take place, or from an increased usage for certain subpopulations, such as bettors, to changing the devices they use to sign in.
Fraud Impact
As we analyse fraud trends and explore their connections to major events, we must remember the volatility of fraud and that any change we perceive could be the natural evolution of the fraud landscape. That said, there are some changes that we observe which suggest a direct relationship with the football.
Also, while fraud reporting is still incomplete for the entire tournament, we can look at the early stages with some level of confidence, considering between three and six weeks have passed, and we would typically expect most fraud to have been reported by now.
Our analysis of aggregated data from the UK shows drops in fraud volumes when England matches took place. The graph below shows the general volumes (blue line), with the match days circled.
Conclusions:
• The first and fourth matches took place on a Sunday, when fraud volumes typically drop significantly compared to the rest of the week. Comparing these is somewhat more difficult.
• There is a kink in the fraud volumes for both the second and third matches, confirming that we do see less fraud happen when matches are on.
• While fraud reporting is incomplete for the latter stages, it would appear the trend holds throughout the rest of the tournament too, with the following observations:
◦ When quarter finals took place July 5 and 6, fraud volumes were significantly lower than previous Fridays and Saturdays. A similar trend is seen for the semi-finals July 9 and 10.
◦ With less fraud seen on matchday, it could be said that fraudsters were watching the football, independent of who was playing.
◦ The day of the final sees a significant drop in fraud volumes. This is to be taken with caution, as more fraud will be reported in the coming weeks, but there is definitely something to look out for here.
Fraud behaviours
Looking at the fraud in more detail, we see that these sessions are much shorter than those observed
on other days: Just 4% of all fraud sessions reported on the day of the final lasted more than 20 minutes, compared to an average of 13% over the last two months.
Part of this could be fuelled by fraudsters executing fewer payments on these sessions.
We see that, of fraud sessions with a payment, there were 42% fewer sessions with three or more payments. Instead, more fraudsters execute just one payment, suggesting reduced patience or available time to execute their illicit activities.
Another change in fraud behaviors relates to the devices (and platforms) used by fraudsters. On average, one third of all fraud in the UK originates from a desktop device (remember: Just 6% of the general population use desktop devices for digital banking, illustrating how fraudsters act quite differently than normal users). On the key match days, we saw a shift toward mobile devices for fraud, with an average of 20% of fraud from desktop devices, suggesting fraudsters embraced mobile banking for extra convenience.
Those that did use a desktop device to execute fraud on the day of the final were three times more likely to paste credentials than throughout the rest of the tournament.
Copa América
At the same time Germany hosted the Euros, the United States hosted Copa América across the pond. Based on our previous analysis of the 2022 World Cup and the popularity of football in Latin America, we expected to see the same decline in digital banking sessions and fraud in the Americas during Copa games that we witnessed in Europe during Euros matches.
As seen below, session volumes did decline, but there were some surprising results.
As shown in our original blog, digital banking activity in Argentina on the day of the 2022 World Cup final (won by Argentina) was just a third of that seen the previous week. Meanwhile, for this year’s Copa América final (again won by the Argentinian side), while there was a decrease, it was just 40% lower, suggesting the regional tournament garnered less interest than the World Cup. This is somewhat understandable, given Argentina were the defending champions, and this could result in less hype.
In Colombia, we see a different story. This year was the first time that the Colombian team reached the final since their win in 2001, which was likely to generate heightened interest. As expected, we saw a significant drop in session volumes in Colombia during the final, with as much as 63% less digital banking activity while the match was played.
Meanwhile, in other countries where football is less popular, we observed no difference in either digital banking sessions or fraud on matchday, as shown below.
Fraud during Copa América
Observing fraud volumes from Argentina and Colombia (this year’s finalists) within the timeframe of the final, we see that in Colombia there was a significant drop in fraud volumes. Meanwhile, in Argentina, volumes were consistent with those seen throughout the tournament.
Also, while 42% of all frauds happen from a desktop device in Colombia, we see that during the final, all frauds came from mobile devices – all in all, similar trends to what we witnessed throughout the European tournament.
With these conclusions, both from this year’s tournaments as well as previous editions, it is important for banks to be aware of how these changes can impact their fraud strategies so that they can make the necessary adjustments to adapt as needed.
