Analytics tools and systems designed to collect and process data are still in the early stages of deployment across the enterprise world. Many IT admins know why they should be using data analysis techniques as they offer the opportunity to increase innovation, efficiency and, ultimately, revenue. But entering the world of prescriptive and predictive analytics can be a daunting task for the uninitiated.
There are three fundamental building blocks that need to be put in place to integrate analytics into business processes and applications, according to Bob Rogers, Chief Data Scientist for Big Data Solutions at Intel.1
- Firstly, businesses need to deploy advanced computing power to enable real-time, in-memory processing of vast amounts of data.
- It’s imperative to allow data scientists to develop analytics models using machine-learning techniques that can extract value from unstructured data.
- Finally, the use of frameworks enables software developers to easily incorporate data science models in the applications they develop.
“It’s a challenging task because at any one time, there are millions of digital transactions flying around that need to be approved or rejected in a split second.”
How HSBC counters fraudulent transactions with real-time data analysis
As one of the largest financial institutions in the world serving over 52 million customers across 74 countries and territories, HSBC is consistently fighting a battle against fraudulent activity.2 It’s a challenging task because at any one time, there are millions of digital transactions flying around that need to be approved or rejected in a split second. To complicate matters, all have varying levels of risk attached to them. For example, a payment request of £2,000 to purchase a HDTV would qualify as a high-risk transaction. Rejecting a legitimate purchase would anger a loyal customer, result in the fee income from the purchase being lost and could even lead to account churn. But if the transaction is fraudulent, approval would mean the customer becomes a victim of crime and the bank is £2,000 out of pocket.
To monitor its customer transactions, HSBC deploys the SAS® Fraud Management solution. This monitors multiple lines of business on a single platform, scanning purchases, payments, fund transfers, and non-monetary transactions in real-time and offering sub-second response times. Instead of using an algorithm to predict customer behaviour, SAS allows HSBC to use raw data to monitor 100 per cent of customer transactions. This unique “signatures” approach captures customer behaviour patterns from every source and evaluates that information every time a transaction is scored, helping understand how customers transact and even how they conduct their overall relationships with the bank.
Predictive analytics revamps Aviva’s modelling strategy
By investing in SAP InfiniteInsight predictive analytics technology, Aviva switched from a strategy of building generic models for all customers to propensity-based models for individual customer groups.3 This allows Aviva to be confident it’s not spamming customers with offers that may not be relevant to them. Instead, the insurance provider sends tailored offers which increase its campaign response rates and increase return on marketing.
“In the past, it has very much been done on what the marketers think might be good. They used gut feel. What we are trying to do is put the evidence into that system and say, because in the past this has happened, in the future we can do something better,” explained Margaret Robins, an analyst at Aviva.
The above examples show how using high-performance data storage and analysis tools can provide valuable insights across virtually any industry. Is your business using data in the most productive way possible to improve internal workflows and customer experiences?
Stay tuned to the Intel IT Center for more on our series of to find out how analytics can transform your workplace
For more information: