Who’s in the Room When the Automation Decision Is Made?

AI’s appeal is easy to understand. A tool that summarizes a meeting, drafts a client communication, or updates a record can reduce repetitive work and return time to advisors and service teams.

But a fast task is not necessarily a better workflow.

Take meeting preparation. A firm may know that employees spend hours assembling materials. Before buying a tool to accelerate that work, it needs to understand why those hours are spent. Are employees retrieving the same information from several systems? Are reports assembled differently because clients have different needs, or because each advisor follows an undocumented routine? Is the delay in preparation itself, or in reviewing information and assigning follow-up work?

Those distinctions matter. A solution that produces an agenda in seconds may still require extensive checking, reformatting, and manual handoffs. Another might remove most of that work. The difference lies partly in how the tool fits the surrounding process.

In wealth management, this is complicated by the value placed on personalized service. An advisor’s professional judgment should not be reduced to a rigid template. Neither should every inherited preference become a permanent requirement for a firm’s technology. Firms need to know which variations benefit clients and which reflect avoidable complexity.

That knowledge often lives in the day-to-day experience of client service, operations, compliance, and advisory employees. Process maps and ticket reports can identify where friction appears. The people handling the work can explain what the reports miss.

Recent research reinforces the importance of operational discipline, even if it does not directly answer who gets consulted on technology decisions.

In a September 2026 study conducted by Cerulli Associates in partnership with Vista Equity Partners, 64% of surveyed RIAs reported that AI had reduced manual and administrative work.[3] Yet half of the 68 participating wealth management firms remained in the lowest tier of a proprietary AI maturity framework, and just 12% reached its highest tier. Leading firms were distinguished by governance, named ownership, training, and measurement of results. The maturity score was developed by Vista, an investor in enterprise software and AI, and is not an independent industry standard.

The same study found that 67% of surveyed RIAs planned to add client service associates over the next two years, alongside 73% planning to hire junior advisors.[3] Automation and the demand for human service capacity are not necessarily moving in opposite directions.

There is a revealing limitation in the study’s methodology. Participation was intentionally restricted to executives, technology leaders, and senior advisors with firm-level visibility into AI strategy; 59% of respondents were C-suite executives.[3] These perspectives are important for evaluating organizational strategy, but they cannot tell us what the employees managing daily workflow exceptions think about the tools their firms are selecting.

An earlier WealthManagement.com survey provides a partial view of how those choices are made. The publication emailed invitations to active website users in late 2025 and received 241 responses. Only 27% of respondents were very or completely satisfied with their firms’ overall use of technology. Thirty-eight percent said firm principals made technology decisions while non-C-suite employees oversaw implementation and support.[4]

That division may be entirely reasonable. Senior leaders are accountable for budgets, security, risk, and enterprise priorities. But neither survey tells us whether the people tasked with implementation were involved in defining the problem before a solution was chosen.

In the research reviewed here, we have figures on who decides and who implements. We have much less direct evidence on who is asked.

Service data is indispensable to finding patterns. A high volume of similar requests may signal a repetitive process worth automating. Longer turnaround times may expose a bottleneck. But those numbers cannot always explain the cause.

A repeated request could reflect a cumbersome process, unclear guidance, poor data, or a genuine difference in what clients need. An operations employee handling the exception may understand that distinction immediately. Without their input, a firm could automate the visible symptom while leaving its cause untouched.

That is why frontline involvement should be a defined part of technology selection, not an invitation to offer feedback after rollout.

Start by mapping the actual workflow and its common exceptions before vendors demonstrate what their products can do. Include the people who complete the work, the advisor who depends on it, and those responsible for compliance and systems integration. Identify which steps require judgment, which are repetitive, and which should disappear rather than be automated.

Then give service and operations employees a formal role in evaluating solutions. A short written assessment should explain what a proposed tool changes for their team, what new review or maintenance it requires, and which exceptions it cannot handle. That assessment should inform the decision alongside cost, security, and strategic fit.

Finally, define success together. Hours saved matter, but so do reduced rework, fewer preventable errors, clearer handoffs, and reliable completion of the full process. The question is whether the firm operates better, not merely whether employees use the product.

That evaluation also needs to extend beyond implementation. Six months after a tool is introduced, someone is still responsible for monitoring its output, managing exceptions, and addressing problems as workflows evolve. Those responsibilities should be considered when estimating the time and resources an automation will actually save.

Consider how this changes the decision itself. Instead of asking an operations team whether it can support a tool after procurement, leadership asks which steps need to change and what evidence would demonstrate improvement. The outcome might be a different vendor, a narrower use case, or a decision to fix an existing process before investing in software. That is not slowing innovation. It is determining whether the proposed innovation solves the right problem.

None of this means every user gets a veto. Some resistance reflects the difficulty of changing familiar habits. Firms must sometimes standardize processes that have become unnecessarily varied. But the distinction between a valuable exception and an avoidable habit is precisely why operational insight belongs in the room.

Wealth management has long emphasized the expertise that advisors bring to client relationships. Its next phase of innovation should take the expertise behind those relationships just as seriously.

The case for involving client service and operations professionals in technology decisions is not simply about recognition. It is about making better decisions with better information. Firms cannot reliably improve work they have not taken the time to understand.