The growth plan
Picture an RIA owner planning next year. They watch a model turn one meeting transcript into a client summary, a recap email, and a set of CRM tasks in a single pass. The plan writes itself: each advisor carries more clients, the next hire waits, and the fee holds.
That plan works if the tool does next year what it does today, at about today’s price. That’s a forecast about a supplier’s compute, and most plans treat it as settled.
Suppliers move in
The suppliers are moving toward this work. In September, Anthropic launched Claude for Financial Advisors, which connects Claude to custodians, portfolio platforms, and CRMs, with skills for meeting prep, post-meeting notes, rebalance explanations, and estate and tax briefs. Others will likely follow. As they do, more of an advisory firm’s output will come from outside the firm.
Who controls capacity
Anthropic’s S-1 shows how a provider thinks about the capacity behind a product like that. It manages compute “as a single, fungible pool” and directs it between workloads “based on where the marginal unit is expected to be most productive at any given time.” Those workloads include research, customer demand, and Anthropic’s own use of Claude. The allocation is the vendor’s call.
And capacity isn’t guaranteed to keep growing on schedule. On November 5, AI stocks had one of their worst days in years as investors worried that opposition to data centers and the midterm results could slow the buildout. If capacity tightens, your firm’s volume is one claim among several on the same pool.
Price and terms
Price and terms carry the same uncertainty. If a provider raises prices or changes terms for its heaviest users, a firm that sized its fees and team to today’s cost absorbs the difference. Volume compounds it. A firm that runs every meeting and every client through the tool uses far more than a firm trying it on a few cases, so the price per task can fall while the monthly bill rises.
Where the cost lands
None of this is a reason to wait. A firm can use these tools well this year without assuming cheap, abundant supply forever. The trouble starts when it builds commitments on that assumption: a fee schedule, a service model promised to clients, a team sized to the output, a book grown to match. Those are monthly obligations, and they don’t flex the way a vendor’s allocation can. If the tool gets scarcer or pricier than planned, the owner pays. Some comes out of margin, some goes to clients as higher fees, and some goes back to people doing work the firm expected to hand off, on a team it chose not to build.
Growth that depends on an assumption you don’t control is a bet on someone else’s future. Make it knowingly.
How Athena builds
This is how I think about it at Athena Advisory, Inc. We build for firms where a few advisors serve many clients, so the core of the work, the planning math and the year-by-year cash-flow projections, runs on our own engine and doesn’t cost more with every client you add. AI helps on top of that foundation, not underneath it.