Can financial services reach the AI summit?
Unless the focus on data changes, the summit of Artificial Intelligence will remain in the far distance for nearly all financial services businesses writes our CEO, Andrew Walsh.
Until recently, data was a word used only by scientists and technologists. Now it’s on the front page.
It’s directly linked to privacy, reputation, commerce, politics and compliance. This year we’ve seen these elements play out publicly and globally with companies such as Facebook.
But data itself isn’t the current focus of concern. Rather individuals are concerned with the protection of it.
It would be easy for businesses to see data as something that needs to be tightly controlled, wherever it is found. The reality is data is a neutral component with benefits only once it is properly harnessed and a danger when protection is ignored.
So while ensuring the protection of data is a necessary part for any business, it’s only the first stepping stone in a successful data strategy — one that also includes business automation and efficiency, and competitive advantage (and profitability, and in some cases, survival).
The link between data and AI
AI is the nirvana of data use but it comes only with large quantities of quality data. Businesses such as Google and Facebook have a distinct data and AI advantage because they have the volume of data that machines require to identify trends and learn. Without data, computers can’t do much, let alone learn.
“While ensuring the protection of data is a necessary part for any business, it’s only the first stepping stone in a successful data strategy.”
The financial services sector generates a lot of data, but many businesses are small to medium in size. Increasing regulation and reputation risks keep the business focus forced towards data protection and compliance (the things that must be done). Only some businesses are more advanced and using data for business efficiency and competitive advantage (the data advantage).
The problem with data
The reality is that most small to medium financial services are in a phase of data transition. From paper to digital and from unstructured to structured.
We recently assisted a client with transitioning a large number of paper records into digital form including scanning, categorising and indexing. That client was a large bank but it’s work being undertaken across many businesses. Businesses are doing this to keep on top of compliance needs, including reporting, or to be better positioned to understand their business through data and anticipate their data future.
Despite progress, the key problems businesses have with data remain around multiple data sources (and multiple systems), data that is inconsistent or overlapping, poor quality data and incomplete data — something that makes data more of a liability than an asset. It is certainly costly. This data landscape in my experience is evident and a core issue to address for any AI ambition.
Why data will be an expected norm
Is a new approach to data needed? Key trends in the delivery and regulation of retail financial services would suggest so.
For example, for digital advice to be easily accessible, it must be componentised — this relies on processes and data points. A digitally-delivered and scaled advice model with the business purpose for cost efficient, focussed, transactional advice units, means success will come from clear propositions, clear processes, and automated calculations and scenarios. But with greater automation involved, there will be greater expectations on data.
As more advice is delivered digitally (whether automated and human-assisted or client-driven) data will be an expected norm.
“Businesses need to look beyond data protection and compliance, to data for business efficiency and competitive advantage.”
Regulators will naturally have even greater data expectations in the future. Signs point to regulators needing data to be even more readily available in the future as they try to stay on top of their roles by using data pools to run their own analytics. Few businesses are prepared for this.
The same applies for businesses who need to see, learn from, and anticipate their own data in real time. Businesses should expect to use data to learn more about their clients and their own process and services improvement, but not forget that data comes from considered processes and systems use to capture, store and analyse.
While data volume will always be advantageous, small and medium businesses will benefit from considering processes and systems to ensure the data they capture is complete, doesn’t overlap, and has quality, so that maximum potential for computers to learn can be established.
AI and the future of financial services
The potential offered by AI as it expands across all aspects of financial services including banking, portfolio management, advice, trading, customer service, underwriting, security, and compliance, is set to solve some significant challenges — and deliver time savings, reduce costs and add value.
There’s little doubt that AI will play an important part in the delivery of financial services in future. But if businesses want a part in that future, the majority need to look beyond data protection and compliance, to data for business efficiency and competitive advantage.
All businesses need data to be smarter: to understand their status at any given time, to be ahead of issues, and to be on top of opportunities. With clear data quality machines can provide broader and faster insights. Each business is in control of its own data quality to help drive this.
This article was first published on iress.com.
