How AI Is Transforming Wealth Management Operations

It’s possible to take a very simple view of a business and split everything into either a cost centre or a revenue centre. The revenue centre usually gets most of the attention and nearly all the praise, after all, it is revenue that pays for everything else to happen.

But the role of a cost centre is incredibly difficult. In simple terms, it must keep itself as small as possible whilst investing appropriately in the people, processes and technology that drive efficiency, allowing as much revenue as possible to become profit.

COO’s will tell you that is as difficult as it sounds and it was harder still when we look back through history and so it is no surprise that businesses have readily embraced technologies capable of transforming productivity. Whether it was machining and steam power during the Industrial Revolution, or the arrival of new technology from the Big Bang of the 1980s onwards, the prize has remained broadly the same. Greater capacity at a lower unit cost.

When technology arrived in the wealth management industry (many years before it was called wealth management) the change was profound. Ledger clerks, manual books of record and post rooms gave way to computer systems and email, rapidly increasing productivity. But, over time, those technological gains began to plateau.

For a while, that plateau was easy to overlook because operational capacity continued to increase. Anyone with a few decades in the industry will recognise where much of that additional capacity came from. Once the first wave of technology had delivered its gains, operations teams took over. These teams were typically the first into the office and the last to leave. They kept everything moving through hard work, accumulated knowledge and an uncanny ability to build complicated workarounds in Excel. At the end of the year it was expected that they would be well paid for the additional effort. But a few things have changed since then.

The Emperor’s New Clothes

Two large global events undermined the traditional back-office model.

The first was the Global Financial Crisis. It put a sizeable dent in industry revenues for several years and inevitably reduced the rewards available to the people doing extra to keep the operational wheels turning. When the extra hours stop being rewarded, the incentive to work them gradually drifts away.

The second blow was COVID. Working late to solve problems as part of a team can be surprisingly enjoyable, doubly so when followed by a quick celebratory drink. Staying late to solve the same problems alone at home, during lockdown, was a far less compelling proposition.

Put the two together and the flexible operational capacity previously provided by a relatively small group of people has all but disappeared. As a result, businesses have found themselves carrying the higher fixed cost of larger teams working more conventional hours. The flexibility, and the reward for going above and beyond, have largely become things of the past.

Historically, this would have been the moment for technology to step in and revolutionise operational capacity. But no such revolution came. In many respects, we went in the opposite direction.

Unsurprisingly, the regulatory response to the Global Financial Crisis was a demand for greater control, rules and rigour. Technology firms, quite reasonably from the perspective of their shareholders, doubled down on that demand.

As a result, financial services technology began to prioritise rules and rigour over true productivity.* Dashboards presented more data, configuration created more rules and workflows turned administration into a series of controlled production lines. All of this reassured compliance and risk teams, but frustrated the front office and, very often, its clients.

The back office was also given less room to interject. The benefit of finding a new way to make a process happen faster became limited, while the personal consequences of trying, and getting it wrong, became severe. The operational know-how that had previously kept the business moving was constrained by technology at precisely the time when we needed efficiency alongside the rules and rigour.**

Much of the investment was deemed successful. New systems, detailed reporting and tightly configured workflows created the desired control. The people who selected and implemented them could demonstrate the results, but they were only ever solving half of the problem.

All too often, the system “out of the box” did not match the processes the business had decided it should follow. This led to customisation. But each customisation dragged the system further away from the best practice processes it had been designed to support. The system became more complicated, changes became harder to make and, steadily, each newly configured process began to slow down.

Where an existing system lacked the flexibility to meet every requirement, a new system was layered on top. If that system did not operate according to the required rules and rigour, it too needed to be configured. Each layer addressed an immediate problem, but added another dependency, another interface and another place where future change could become stuck.

The “clothed” belief that investing in technology would drive efficiency held the industry’s attention. The “naked” reality was that much of the technology was being designed and configured for a different purpose: to provide control, consistency and evidence that the process had been followed.

Over time, the complexity created by those decisions has become technical debt. Its age makes it less stable; its configuration makes it harder to change. Technology did not fail; it achieved most of what it was designed to achieve. The problem is the changes described as operational transformation were in reality more complicated processes.

* We purposefully preface productivity with the word “true” because technology has delivered productivity gains. However, those gains have been swimming against a regulatory tide that has offset much of their impact. Technology has helped ensure that things are not as inefficient as they could have been, but that is different from fundamentally improving operational capacity.

** Two things can, of course, be true at the same time. Greater robustness and rigour were necessary lessons from the Global Financial Crisis. But the resulting constraints on back-office know-how, combined with underinvestment in operational efficiency, have also reduced the industry to something of a slow plod.

The Operating Strategy Trade-Off

Outsourcing has traditionally offered wealth managers the opportunity to convert a largely fixed cost base into a more variable one, together with the expectation that the provider will invest continually in improving the service.

That benefit is real, but variable cost is not automatically lower cost and outsourced investment does not always align with a wealth manager’s priorities. Equally, firms retaining an in-house operating model gain greater control, but must fund future investment themselves.

For many years both models have proven viable. Firms have regularly moved from one approach to the other, and in some cases back again.

But these are not normal times. AI has the potential to be one of the most transformative technologies the industry has experienced since the arrival of computerisation.

Rather than choosing between wholly outsourced or fully insourced operations, a different model is likely to emerge. AI will increasingly sit above existing operating models, automating activities wherever it can create more value. That may mean replacing elements of an outsourced service, augmenting internal operations or removing the need for some systems entirely.

The result is not the wholesale replacement of one model with another. It is a constantly shifting boundary between what the wealth manager performs, what providers perform and what AI can increasingly do better than either.

This makes control more important, not less. Wealth managers do not need to own every system or perform every process, but they do need sufficient control over their data, architecture and commercial arrangements to move that boundary as technology evolves.

Before those choices can be made, firms must understand what their current operating environment actually allows. The first step in transforming an operating model is securing access to the data needed to understand and change it.

The Three Anchors

Before making any operating model decision, a wealth manager must be clear about what cannot be compromised:

• Client assets must be safe.
• Data must be clean and accurate.
• Clients’ access to their data must be instant.

These are the anchors of any future operating model and wealth management proposition. Technology will change, providers will change and processes will change, but these obligations will not. Every decision about outsourcing, architecture, automation or AI should ultimately be judged against them.

Investing Around the Anchors

Once those anchors are established, the question shifts from what must remain stable to where firms should invest, innovate and evolve.

The traditional view was that firms should invest in systems expected to serve them for ten or fifteen years. Choose carefully. Implement thoroughly. Build for longevity. That approach made sense when replacement was slow, integration was difficult and switching costs were prohibitive.

Today, the principles remain long term, but the technology supporting them does not need to be. Books and records, controls, data definitions and ownership responsibilities require stability, but the applications, workflows and intelligence used to operate them should continually compete to remain relevant.

AI accelerates that shift. Some core technologies linked to asset safekeeping and trusted data may remain in place for another decade, but firms no longer need to anchor themselves to every surrounding system. Instead, they can invest in capabilities that improve processes across the back office, middle office, front office and ultimately the client experience.

Individual agents should operate as part of an agentic mesh, connected to the same trusted data and governed through a common control plane. The three anchors provide the purpose and boundaries for that model: no agent should compromise the safety of client assets, the accuracy of data or the client’s ability to access it.

Changes to the Cost Model

There are some things clients care deeply about but have little interest in paying for explicitly. Most of these sit within the cost centre and, in wealth management, two of the most important are the anchors of custody and data. Both are essential. Failure in either can cause enormous harm, but neither is usually why a client chooses one wealth manager over another.

This makes custody and core data services commodities, although they are commodities with exceptionally high standards. They must be dependable, accurate and secure, but clients reasonably expect them to be included within the service rather than presented as sources of additional value.

Providers of commodity services are accustomed to continual pressure on their fees and must evolve to remain viable. Custody providers have invested in scale, technology and standardisation, with many expanding into data and related services. Custody fees have already experienced considerable compression. Data, at least for now, is holding the line.

AI and automation are likely to increase this pressure, both for providers and for wealth managers themselves. The revenue implications are equally significant and are themes we will continue to explore through future thought leadership. The cost of the back office may not fall immediately, particularly while firms invest in new capability and operate existing processes in parallel. But its capacity should increase substantially and, as it does, the cost of each unit of activity should decline.

This changes the shape of the cost model. Traditional back offices carry substantial fixed costs in people, systems, controls and management. Capacity must often be installed before it is required, and additional growth eventually triggers another step-up in expenditure. Outsourcing was the historic option for making some of those costs more variable. AI presents a new option: increasing the capacity of the existing cost base before the next step-up is required.

Neither approach is automatically beneficial. A variable charge can become expensive when assets or activity grow, and poorly controlled AI can add complexity rather than remove it. The objective is therefore not simply to replace fixed costs with variable costs or people with technology. It is to create a cost base that responds intelligently to activity while allowing capacity to grow faster than expenditure.

This presents wealth managers investing in AI with two choices:

• Harvest the efficiency: Use the additional capacity primarily to reduce headcount, supplier expenditure or other operating costs.
• Use the capacity for growth: Support more clients and advisers, launch new propositions, improve service and absorb greater complexity without increasing expenditure at the same rate.

Some efficiency will inevitably be harvested, particularly where AI removes activity that no longer needs to be performed. But treating cost reduction as the destination is a trap.

The real opportunity is to reinvest the majority of the capacity created into growth, allowing the business to increase revenue faster than its cost base.

Capacity Creates Choice

For several decades, the wealth management industry relied on operations teams to create the capacity its systems could not. Experienced people connected fragmented processes, interpreted incomplete information, solved exceptions and absorbed peaks in demand. Their knowledge, effort and ingenuity allowed businesses to grow beyond the apparent limits of their technology.

That model was remarkably effective, but it was neither scalable nor permanent. The financial settlement that rewarded discretionary effort weakened, the working environment changed and successive waves of technology prioritised control over genuine productivity. The industry became safer and more rigorous, but not proportionately more capable.

AI creates the possibility of changing that equation. Its greatest potential is not a better dashboard, a faster workflow or a collection of isolated automations. It is an agentic mesh capable of coordinating activity across systems, providers and processes, while escalating consequential decisions to people with the judgement and authority to make them.

• The Three Anchors remain unchanged: Client assets must be safe
• Data must be clean and accurate
• Clients’ access to their data must be instant

Everything else should be open to challenge.

Wealth managers should use outsourcing where a provider’s scale, specialist capability and pooled investment produce a better outcome. But they should not surrender access to their data, control of their differentiation or the ability to transform. They should demand resilience from the core of the operating model, while ensuring that the technology and intelligence around it remain open, modular and replaceable.

They must also be honest about what they are trying to achieve. Applying AI to individual tasks may produce worthwhile incremental improvements, but it should not automatically be described as transformation. Transformation occurs when connected capabilities change the capacity, economics and addressable market of the business.

Most importantly, wealth managers should decide what they intend to do with the capacity created. Some will use it defensively to reduce costs and protect margins, but it is impossible to shrink to greatness, even with something as powerful as AI. Others will use it as firms once used their back offices: to support more advisers, serve more clients and improve the experience they provide, while keeping a controlled grip on costs.

The back office should no longer be viewed simply as the part of the business that must be kept as small as possible. Properly designed, it can return to its position as a provider of operating leverage, protecting the obligations on which trust depends while allowing revenue to grow faster than cost.

In that respect, the objective has not changed. The operations teams of the 1980s, 1990s and early 2000s used knowledge, effort and ingenuity to power the growth of financial services. The opportunity now is to capture those same qualities in an operating model that is controlled, measurable and scalable.

The next generation of wealth managers will not win because they have the lowest-cost back office. They will win because their back office creates the greatest safe capacity for growth.