Electronic transactions are the largest, fastest growing subset of “Big Data” within financial environments. Although the complexity and volume of this data may seem overwhelming, many channel managers, marketers and operations teams realise that - when harnessed and leveraged correctly - transaction data also represents a predictive, “always on” feed into how the organisation is performing, how services are being delivered and how happy banking customers really are.
According to Bob Meara, Senior Analyst at Celent:
“62% of financial institutions in a recent Celent survey strongly believe that customer analytics offers significant competitive advantages and 53% strongly feel they need a granular, holistic and forward-looking view of customers to be competitive. One key to understanding customers and improving banking channel efficiencies lies in making rich transaction data accessible for actionable customer analytics.”
DIAGRAM: Who Uses Transaction Data
The challenges of transaction data mining
But mining transaction data can be a costly and time consuming task. In recent years, the ability to collect, correlate and extract “actionable” transaction data has become more challenging for reasons such as:
- The massive, growing volumes and diversity of electronic banking and payments transactions
- The omni-channel convergence of ATM, POS, Branch, Mobile and Internet Banking channels
- The increasing disparities between how banking technology platforms work, including virtual, cloud-based, mobile and Agile applications
- The growing number of ways in which transactions get processed through banking services, fraud applications and payment networks
- The merger activity and globalisation focus of many financial organisations
The good news is that fresh thinking and new approaches have made it a whole lot easier to overcome these challenges. Recent technology advancements in real-time transaction data collection, storage and vertically focused visualisation applications will help position your organisation to win.
Introducing new ways of collecting and storing transaction “Big Data”
For today’s channel managers, marketers and operations teams to meet customer service expectations, they must be able to discover performance issues and respond to changing customer behaviours extremely quickly.
This is why it has become increasingly important to collect and store transaction data in a centralised location, across all your moving network parts and channels. Centralised data collection means that you don’t have to go searching for it when you need it the most.
When you use transaction data for customer analytics, this data will ideally be captured across an entire “end-to-end” transaction path. Why, you may ask? It’s because your customer analytics will only be as complete as your data.
Each structured individual customer transaction record will contain a wealth of information, including:
- Terminal ID and geographic location information
- Transaction approval and event completion status
- Transaction types and dollar amounts
- Card types and host authorisation responses
- Response timing and failure information for each “moving part” on the transaction path
Traditionally, channel business managers and marketers have lacked on-demand transaction data access, and have relied on their operations team to spend many expensive, time consuming hours piecing together fragmented data collected from multiple sources – many of which are unreliable. These data collection and storage issues surrounding transaction data have also impacted the ability of retail banks to produce dynamic dashboards and customer analytics reports without long lead times.
Thankfully, the days of deep dependency on IT for customer intelligence are now over. Many of today’s retail banks now depend on software to do their heavy data lifting - not their operations team.
Today, there are real-time transaction monitoring solutions available, that are built to handle high-volume electronic transaction environments. They easily monitor any and all types of electronic customer interactions happening within complex multi-channel retail banking networks. Protocol and platform agnostic, these software platforms capture and correlate real-time consumer transaction data in a secure, light-weight fashion, enabling financial institutions to overcome past data mining limitations and avoid high cost deployments, manual data collection and time consuming correlation tasks.
Warehousing this data within an environment such as a Hadoop data store means that you will have the ability to easily scale to a cluster without costly, additional licensing required. This data, along with any complimentary third party data you wish to store, can be extracted from Hadoop using a friendly, self-serve analytics interface such as Tableau, or simple SQL queries – no extra code writing required.
Using transaction Big Data to drive customer banking engagement
Now that you have the centralised collection and storage of your transaction data figured out, it is important to think about how you want to use that data. To address specific customer use cases or quickly execute data-driven actions, the front end customer analytics solution being used for data visualisation should meet the following vertically-focused criteria:
Contains multiple preconfigured dashboards designed to expedite the analysis of customer engagement patterns and conduct predictive analytics
Can embed transaction data and other feeds for specific use cases
Contains reports that can be run on-demand and easily understood by line of business managers and marketers
There are more and more examples of Big Data and analytics vendors tailoring their offerings to specific verticals, offering self-serve solutions that meet highly specific reporting and dashboard needs around channels such as ATM, POS, Mobile, Branch or Internet banking. This means the days of deep dependency on IT for the creation of intense analytics querying, reports and dashboards are also over.
Self-serve, analytics software is an easy way to see your wealth of customer transaction data – ready to be analysed any time you need it. With on-demand access to rich records of every consumer transaction, channel business managers, marketers and operations teams can apply their rich domain expertise and predictive modelling to understand customer behaviors throughout any banking channel, such as:
- Who uses which ATMs or POS terminals?
- What kind of interactions are customers having at my Internet Banking or Mobile Banking application, and when?
- What quality of service are customers experiencing?
- How profitable is this customer or group of customers?
- How did that latest outage affect channel profitability?
You can also ingest complimentary third party data sources such as BIN lists, ATM configuration data, competitor location information and customer demographics for more enriched data visualisation and decision making.
Taming “Big Data” transaction complexity
Whether you are interested in making better ATM or POS placement decisions, improving cash inventory and replenishment schedules, simplifying reporting or running ad hoc queries on outage impacts, transaction data is your goldmine.
The key to extracting actionable customer analytics from transaction “Big Data” is exploring new ways to manage and correlate high volumes of diverse transaction data so problem identification and customer engagement analysis becomes simpler, faster, and more accurate.
New technology advancements in real-time transaction data collection, storage and vertically focused visualisation applications are designed to be activity centric: providing relevant, timely, and actionable information that is well suited to the event-driven nature of today’s omni-channel banking environments.
By Stacey Gorkoff, VP of Marketing at INETCO