The world’s stewards of capital have realized that they can’t do without artificial intelligence (AI), the technology that generates insights from mountains of data to drive enterprise value.
But not everyone is moving at the same speed. A gulf is opening up between the leaders, who have taken the plunge and industrialized and scaled AI across their organization, and those who are merely dipping their toes in the water. Some asset managers are now reaping up to 300 basis points (bps) of alpha through the use of a full, comprehensive range of AI solutions, according to Accenture ACN -0.6% research. Meanwhile, the toe-dippers are just squeezing a few basis points per individual use case, if they’re seeing any benefit at all.
It is tempting to apply Leo Tolstoy’s theory about happy and unhappy families when explaining what AI leaders get right, and what the stragglers get wrong. The leaders all seem to avoid the major pitfalls, but a single flaw in an AI strategy – whether that is technical, cultural or governance-related – is often enough to doom a firm to failure. But one thing is certain: If asset and wealth managers want to establish themselves as AI leaders, they should call time on their science experiments with the technology. And there are plenty of firms marooned at the experimenting stage: More than seven in ten(72%) wealth managers Accenture surveyed in the fall say they are running siloed proof of concept projects or deploying AI in targeted business groups.
But capital markets firms taking the plunge can study the governance and culture of AI leaders. The successful firms avoid AI strategies that are a patchwork of small-scale pilots, prototypes and trials because they know this piecemeal approach will seldom bear fruit. Instead, they set out a bold vision for adopting AI at scale across their organizations.
This does not mean iron-hand central governance structures and micromanagement, but the opposite: Governing and stepping away. Companies should decide on the overall direction of the travel on AI and then delegate to business and portfolio managers. This allows domain experts to work side-by-side with data scientists on data mining, signal generation, intelligent optimization, and predictive techniques. If firms still count on a small team of scientists who are responsible for an entire company’s AI strategy, they’re headed in the wrong direction.
As they delegate, boards and top executives should underline the importance of data and data management to the business, as there’s a clear correlation between AI success and the volume and breadth of data. Nearly half of asset managers in our research have caught on, and are now using unstructured or alternative data, whether it’s satellite imagery, weather patterns or transaction data, Over time, firms may not even draw the distinction, because there’s nothing ‘alternative’ about highly predictive data that can help an ML model generate better insights. But these data mountains also require deft management, and only one in two asset managers surveyed are taking the time to normalize and standardize data ingestion.
There’s no doubt that AI can transform the middle and back office, but as they draw up their bold plans for the technology, firms should not neglect the place where real enterprise value lives: The customer experience. Many (65%) wealth managers believe AI can create the most long-term value in the middle or back office, while only 35% think it’s in improving the client experience and engagement.
But these issues pale in comparison to the inability of some firms to exit the AI experimentation phase and enter the era of large-scale production. If asset and wealth managers don’t move fast to achieve scale with good data, governance and culture, they will pay a hefty price.
This article originally appeared on Forbes.
