Dale Kim senior director of technical solutions at Hazelcast discusses how in-memory computing and AI are the key to up-selling at scale in the financial services sector
Banks are constantly looking for opportunities to up- or cross-sell products to customers. Increasing product penetration from 2.5 products to 4 products per customer can add millions to the bottom line and it is estimated to be 5 – 10 times cheaper to up- or cross-sell to an existing customer than to acquire a new one. Combining in-memory computing with AI opens up new opportunities to do so.
The need for real-time insight
When it comes to engaging customers in up- or cross-selling conversations, timing is everything. Customers are far more likely to be receptive to an approach when they are already interacting with the bank – online, via the telephone, or in branch. But for this to happen, the bank needs a real-time, comprehensive and contextualised view of each customer, including the accounts they hold, their transaction history and much more.
That’s where things get challenging. At any given moment millions of customers might be interacting with different parts of the bank’s systems via different devices in different locations. New data will be streaming in from computers, smartphones, customer support lines and other sources.
In the past, transactional data was managed as a batch process and analysed periodically – often hourly or daily depending on the data involved, the bank’s own policies and any regulatory and compliance requirements. For real-time insight, transactional data needs to be executed in less than the time it takes to blink. Fortunately that’s become possible through stream processing and in-memory computing.
Stream processing refers to real-time management of data entering a banking system at high speed and volume, usually from a broad range of sources. The data is wholly or partially processed and contextualised before entering an in-memory data grid where historical context can be applied in microseconds to improve the probability of a successful up- or cross-sell.
Combining AI with in-memory computing enables upselling at scale
Where this gets interesting is when all this real-time data is used to power machine learning (ML) and artificial intelligence (AI) systems – whereby the customer thinks they are talking to a human being but are in fact speaking to a machine. As well as improving the probability of a successful up- or cross-sell, the machine may arguably be able to provide a better, more ‘intuitive’, customer experience than a human could. The main advantage of tying an in-memory solution to an AI solution, however, is volume; millions of simultaneous customers calling in can strain even the most significant call centre, but not be noticeable to an AI-powered chatbots.
It should come as little surprise therefore that industry analyst Gartner predicts that by 2022, a massive 70 per cent of customer interactions (in banking and otherwise) will involve emerging technologies such as ML applications, chatbots and mobile messaging – up from 15 per cent in 2018.
The above use case along with others such as more accurate fraud detection in payment processing and a reduced risk of false positives undermining customer experience are what has driven the success of combined AI and in-memory computing technologies within the financial services community to date. There is little doubt that it will continue to do so.
