Banks are under pressure
Like lots of organisations, banks are feeling the pressure right now. Branches (if they are open) and call centres have experienced significantly large volumes, just as the lockdown is taking its toll. Halifax, part of Lloyds Banking group, recently closed a call centre in Northern Ireland and 1,000 staff were sent home. Fear of job loss and the subsequent inability to pay off mortgages or loans is causing serious stress for customers, which, in turn, is putting pressure on both front and back office. Many banks continue to rely on the heroics of their colleagues to orchestrate very different types of applications in order to complete what looks like a simple task – such as changing a direct debit date or enquiring about a transaction. 

To deliver efficiency banks have rightly chosen to focus on intelligent automation (including, for example, AI, analytics, machine learning and RPA). Accenture research estimates that banks that invest in AI and human-machine collaboration at the same rate as top-performing businesses could boost their revenue by an average 34% and their employment by 14% by 2022. Contrary to popular belief, staff roles will increase. But this will require staff training and reskilling for new roles within the bank. Not every teller will become a data scientist, but it will be more cost-effective to do this than recruit expensive consultants or contractors who will leave the business after a short-term contract. 

This approach will also free up staff to have more advice-driven discussions around complex products that lead to greater customer satisfaction and retention. Morgan Stanley is supporting its financial advisor network with AI agents. The agents continually interact with their human co-workers to proactively recommend a range of options for their clients. This helps the advisors to contact clients at the right time with more relevant advice.

People and machines will collaborate
Banks have started using automation in areas such as KYC and AML. Predictive models are being used to prevent high-value customer churn and identify where to cross- and up-sell through ‘next best action’ analysis.  Chatbots are common now. Bank of America’s Erica, completed over 50 million client requests in its first year, including checking payments and balances and monitoring FICO scores. TSB, a UK bank, recently added a 'smart agent' function to its site, enabling customers to ask a chatbot about measures the bank is taking during the pandemic. Banks are digitizing processes to eliminate manual input (and error) where possible and speed up business outcomes. Part of this approach also involves moving to a cloud platform to become more responsive to customer demands. 

With Microsoft and Accenture we digitized a paper-based, labor-intensive loan-approval process for a Brazilian bank, who were the market leader. Customers now use their computers or mobile phones to upload documents and link credit scores for assessment and approval. The new digital platform enabled the bank to implement automated, risk-based pricing, resulting in 25% savings for underwriting and processing. Pre-check and document handling time was reduced from six days to a few minutes. Revenue increased by 40%. Customer satisfaction increased by 35%. Loan generation increased by $400 million.

Improving automation
Here are some suggestions where banks can improve their use of intelligent automation:

1. Don’t forget the Middle Office. AI is being implemented by banks within Middle Office functions to stop payments fraud and improve the process for AML and KYC regulatory checks. It is estimated that the savings here ($217 billion) are as good as those implemented in the Front Office through biometrics and personal insights ($199 billion).
2. Develop conversational AI. Automation need not be a faceless impersonal journey. Indeed, to make it successful it needs to be compelling and personal. Banks need to look at new skills to design and develop more engaging experiences designed around chatbots and digital assistants. Microsoft uses a team including a poet, a novelist, and a playwright to develop Cortana’s personality. Computers understand text; people understand subtext, that is, what’s left unsaid in a conversation – and conversational AI needs that skill too.
3. Focus on data. AI and automation are only as good as the data that feeds it. Most of the time the data comes from disparate databases with little in the way of common classification. Banks need to focus on the data, cleaning it where necessary and making it available to the services that drive the models. Data should be augmented with structured and unstructured data to deliver the most complete solutions.
4. Advice is key. A 2018 Aite study showed that 79% of 22-to-34-year-old consumers and 77% of 35-to-49-year-old consumers indicated that they are moderately to extremely interested in using a virtual financial wellness coach. Ally Bank used virtual reality to create a digital version of the Monopoly board game as a means to improve financial literacy. Other examples of financial advice include Digit, Mint and Multiply (which won FCA approval). 
5. Invest in people. Despite the current pressure to reduce headcount, employees are the key to the way ahead. People are not only keen to thrive in an AI-driven environment, they are also eager to acquire new skills to make this happen. However, Accenture research found that only 3% of bank executives plan to significantly increase investment in training in the next three years. 

Banks have a major opportunity to emerge from the current crisis with a more responsive business and a lower cost base. Accelerating the drive to automation is a powerful way to do so.  

#Coronakindness
With each of our current blogs we’re sharing a good news story …
Spain's Caixa Bank is helping retailers hit by the pandemic by enabling them to sell over social networks and messaging applications. Because their physical outlets are closed, small businesses are now rushing to get online operations up and running. Caixa Bank has launched ‘Social Commerce’, which lets retailers manage online purchases directly from their profiles on social media.

Customers receive offers via the business' social profile and messaging apps. If they wish to purchase, they can do so through a direct link to the payment instructions. The retailer receives the information on the sale and can then manage shipment. Caixa Bank has also launched ‘PayGold’, which lets firms receive payments online - by email or text message - without the need for a website. Caixa Bank is also offering discounts on POS terminal fees to firms affected by the pandemic and has developed a financing package to help SMEs and the self-employed.