AI forces the data-warehousing question at RIAs

At a typical advisory firm, one client file lives in four places: the CRM, the portfolio accounting system, the financial planning software, and a document drive. Each system was built to do one job well, and each holds a piece of the client. AI is now testing whether that arrangement can last.

InvestmentNews this week traced the split to the systems' origins. The CRM grew out of contact management, back in the rolodex era. Portfolio management came from portfolio accounting, which supported performance reporting long before billing and trading. Financial planning software came from its own specialized planning function. Document management arrived with the conversion of paper files over the past 25 years. Together they make a set of mature silos: good at their core jobs, bad at sharing.

The weakness shows up every day. The same facts get maintained in several systems. Servicing a client means checking several screens, and any expansion of workflow demands integration across systems, with many vendors in between. That integration is hard, and it gets harder with every additional software provider.

AI has made that difficulty urgent. To build knowledge bases, and especially the agentic layer that takes actions on a user's behalf, systems need all of the data organized cleanly in one place. InvestmentNews reports that firms betting on an AI-enabled future, one that supports advisors and may replace some back- and mid-office staff, now face a gathering need to centralize and warehouse it.

The need is real; the payoff is not automatic. A warehouse is an infrastructure project, not a software upgrade. It demands data governance decisions, schema design, and continuous synchronization with the systems of record. Costs arrive immediately; benefits come later, and only if the warehouse is fed clean data. A warehouse that simply copies the silos into one database does not solve the maintenance burden. It moves it.

For an RIA principal, the practical question is what the warehouse becomes. Today each system is the system of record for its own slice of the client. A well-built warehouse could become the authoritative file, and that shifts the balance of power among software vendors. It would be the opposite of the world the silos were built in, where each vendor could count on being the single source for its function.

The same fragmentation that pushes firms toward a warehouse also makes one hard to trust. CRM data reflects its contact-management roots. Custodian feeds arrive in whatever format the clearing firm dictates. Planning software has its own assumptions about taxes and goals. An agentic AI acting on contradictory records is a liability in client meetings, not a labor-saving device.

The organizing problem is drawing attention outside the advisory stack, too. PWD reported this week that Edward Jones took a minority stake in Quicken, the consumer app that promises to hold a household's financial life in one place. The two developments are different in kind, but both reflect the industry's convergence on the business of organizing data, because so much of an advisor's value has become organizing rather than recommending.

The firms that treat the warehouse as a strategic asset, with an executive who owns data quality rather than just a procurement team, will likely see AI's benefits first. The rest will have spent seven figures on a large, expensive replica of the same mess.