Real-time analytics allow us to take historical data and apply it to things that are happening right now. It's a mixture of traditional data analytics and insight into what's happening inside your payment system. By combining data about how your system is working with ad hoc reports, we create a better picture of what's going on.
For example, it's possible to take a glimpse at the normal pattern of transactions flowing through your system and then see if something is happening that's not normal for this time of the day (or day of the week). From there, we can provide customized alerts that let you know that you might want to pay attention to the issue before it becomes a bigger problem.
Detecting Normal Patterns
Historical data can help to detect patterns that are typical with payment systems. At the start of the business day, for example, card payment systems might show a slight spike in transactions. The volume will grow steadily over time until it peaks at midday and eases off towards the end of the workday. There may also be a small peak near closing hours while everyone is going to the ATM to withdraw money.
This is a relatively simple example of a pattern that's easy to detect. The power of realtime analytics comes from finding correlations between different points of data. If you can pinpoint times of day where there are bursts in transaction decline rates, that could sometimes be an indicator of fraud attempts. You might not be able to find out for sure unless you know the patterns of your system and look at different metrics together. It might be a combination of transactions, time of day, and opening/closing times of particular merchants. The more you know the patterns, the more you can detect things that aren't normal.
Businesses clearly understand the importance of collecting and analyzing data. There are many tools available, but it really comes down to making sure you're extracting the relevant information that suits your environment. From there, you can use that information operationally to improve the performance of your system. It helps to know your payment environment—that way you can tie the historical data back into the real-time transaction flow. You'll be able to detect abnormalities while they're happening rather than having to wait until you run analytics at the end of the day.
In part two of this blog post I'll discuss how historical data can complement real-time analytics to make meaningful predictions.