Analytics a game changer for ATM Banking?
By : Flytxt
Banks have been on their journey in digitalising their operations and bringing customers to the centre of business activities. They have realised that big data analytics is a key technology enabler to drive their aspirations in both these directions. Analytics can empower channel managers and omni-channel banking executives with the game-changing knowledge of where, when and how customers interact with their bank or financial institution to differentiate their offerings and personalise banking experience.
However there is a third angle to it too, analytics can also help them in improving their operational efficiency. It can bring in better prediction and intelligence in their current operational processes and also their distribution of payment channels like the ATM network (Automated Teller Machine) whose usage has grown exponentially over the past few years.
Convenience triggers knowledge
ATMs as a payment channel emerged in 1970s and became popular in 1980s. Today, ATMs are not merely a payment channel. They function as a mini bank, especially in rural areas, with facilities for customers like opening an account. Some of the biggest advantages of the ATMs are their ease of use, increased functionality and proximity to high foot and vehicle traffic areas. In India itself, you have close to 2 lakh ATMs. Hence, banks can potentially infer massive nuggets of business intelligence from massive volumes of incoming ATM transaction data spread across locations.
ATM transactions are a goldmine of information that can help find answers to questions related to both customer experience and operational efficiency like – who uses them, which ATM locations they use, how frequently, when do they use them, what kind of transaction, total amount they transact, how are various cash points, networks and applications responding, etc.
Banking on Big Data
With ATM transaction data, metadata and other possible data sources, big data analytics can easily help banks to provide:
- a single view into customer interactions on ATM network and the rendered services performances
- consolidate ATM data like currency amounts, card types, response codes, terminal IDs, transaction types and even operational procedures like cash refill, maintenance shut down timings etc.
- periodic transaction statistics such as cash deposit or withdrawal totals, the number of transactions by service type and the number of “on us” versus “off us” transactions
- correlate meta data (i.e. ATM terminal ID, geographic location, transaction status), application message data (i.e. transaction type, time, dollar amounts), response code / error information and network level TCP acknowledgement information into a one-stop hierarchical view for each consumer interaction such as your most popular locations, times of peak usage, cash levels, and the ATMs responsible for your most profitable transactions.
If you are responsible for ATM availability, rich deposits of troubleshooting intelligence can be collected from switch response codes, network and application performance data—intelligence that can alert you to failing ATMs, and more importantly, provide the reasons why they are failing.
One can build a self-serve analytical application that shows how customers use the ATM channel, gain visual insights, measures to deliver more value per transactions (ex: popping up automatically the frequently withdrawn amount) and drive operational efficiency through better placement and service offerings.
Location analytics and more leads to a new wave of opportunities
Using advanced models and data driven decisions, one can also quickly spot the top business value customers, most profitable locations, top violators (my customers transacting in another bank channel), most used localities, market dominance, cost benefit analysis of provisioning a new ATM vs. using the competitor’s ATM machine, ATM queues, wait times, predict cash-out times. This data can also be used to display information such as current available cash positions for each ATM, best replenishment transaction days, times and amounts, and how many customer transactions failed due to a network outage, bottleneck or application issue.
Based on prior transaction behavior, the models can assess customers’ withdrawal patterns across the entire network for each machine. They can also forecast, with a fair degree of certainty, the future transactions frequency, type, amount, transaction time, wait time etc. Specifically in case of withdrawal, using these forecasts, banks can achieve better operational efficiency and service excellence by generating an optimised replenishment schedule at non-peak times for the entire network to achieve minimum cash-outs and number of trips while being operationally realistic and robust.
One can analyse customer transactions and their demographics at big data scale to build a 360 degree customer view and persona to derive actionable insights reflecting their usage behaviour, needs, and preferences. This will then get mapped according to the top three most relevant offers and will be tracked till its fulfillment.
Unlocking the big potential
Regardless of the channel used, every transaction, starting at the customer, goes through various different network, third parties, different servers, authorisations – this entire transaction lifecycle carries a lot of information about the network, the channel, the customer and the service. Big Data Analytics on these transactions can extract multitude of stories that can bring the financial institutions closer to a clearer understanding of what customers want and how it can be delivered in the most cost effective way possible.
Thus, with the right advanced analytics models, banks and other financial institutions can move away from reactive measures to a more robust preemptive model while maintaining and operationalising ATM networks. This will not only reduce the occurrence of undesirable situations, but also significantly improve the overall customer experience at a much reduced operational cost.