Payment Enrichment: The Key to Precise Transaction Data

Raw, unstructured transaction data are the cause of this problem. When you pay for something in a store, at a gas station, or at a bar, this doesn’t necessarily mean that that the payment will appear on your card statement under the actual name of the point of sale. The data instead come in many different formats, making it hard to tell individual merchants apart.

This leads to a number of problems, specifically the following ones:

• Customers cannot keep track of what they spend their money on.
• The lack of traceability makes it harder to draw up a personal budget.
  
• The lack of transparency increases the risk of fraud.
  
• Unstructured transaction data make it harder for banks to design intuitive user interfaces.
  
• Personalized offerings such as a categorization of transaction data in e-banking feeds cannot be optimally provided based on raw transaction data.
  
  

Banks face increased expenses as a result of more incoming queries from customers.All of that complicates data sourcing, and in the worst case, it results in financial institutions learning of corporate actions too late or even missing them altogether. That, in turn, can result in financial losses, but also reputational harm. The same goes for companies that carry out corporate actions. Failure to adequately inform investors about corporate actions is detrimental to a company’s reputation.

A solution to this problem already exists. It’s called data enrichment, a process by which additional information is added to existing datasets to enhance their usefulness. One either adds on external data or refines already existing data. The objective of data enrichment is to obtain more precise and more extensive data. That enables better analytics or even makes it possible to utilize the data in artificial intelligence applications. Data quality is crucial for the latter.

Businesses frequently use data enrichment to optimize customer or product data, for instance, or like in our example case, to enhance transaction data. There’s a separate name for that: payment enrichment.

Payment enrichment is a subcategory of data enrichment. Payment enrichment optimizes transaction data, solving exactly the problem described at the outset of this blog post. Payment enrichment makes payments traceable and transparent. But how does it work? Let’s take a look at the following picture: 

It all starts with raw transaction data because without payment enrichment, an e-banking or payment app user interface would look something like the image on the left hand side. Rather cluttered and hard to read, isn’t it? This example illustrates why raw transaction data are a problem. This is where payment enrichment comes into play. It makes the data intelligible. If we take payment enrichment services from SIX as an example, the user interface would then look like the on the right. But how do you get from raw transaction data to an attractive user interface? Put in simple terms, out of raw unstructured data you extract structured data that are easy for people to understand and simple for computers to process.

Sticking with the SIX Payment Enrichment Services example, each payment gets enriched with the following data:

• Transaction category
• Intelligible name (“pretty name”) of merchant
• Merchant’s logo
• Merchant’s contact details
• Merchant’s location

Several different approaches are combined to enrich the data: data editing algorithms, pattern recognition, and editing by humans.

Payment enrichment helps to render transactions understandable. It makes payments more transparent both for banks and their customers, bringing a variety of benefits. Payment enrichment helps, for instance, to detect fraud because it enables banks to better understand transaction data and thus enhances their ability to spot anomalies. Their customers benefit by receiving clearly understandable information and personalized offerings. The better intelligibility of the data also makes it possible to conduct more targeted analytics.

Since the data become structured, they are ideally suited also for the deployment of artificial intelligence or machine learning. And finally, payment enrichment results in cost savings for banks. Readily understandable transactions, for example, reduce the number of call center queries made to clear up the identity of payments.