SOA Production: From 12 Hours to 3 - A Financial Planning AI Win

Case study: AI-assisted Statement of Advice generation for a North Sydney financial planning practice. 75% time reduction, 31% AUM growth.

The Challenge

Four financial advisors. Two support staff. A North Sydney financial planning practice with a solid client base and good reputations in their market.

But a problem that looked like success until you looked closer.

Each Statement of Advice took [PLACEHOLDER: 12 hours]] to produce.

A Statement of Advice—or SOA—is non-negotiable in financial planning. It's the regulatory document that shows the client what you've recommended, why you've recommended it, and how it aligns with their goals and circumstances. It's essential. It's also [PLACEHOLDER] pages of mostly structured information: client details, goals, current financial position, recommended changes, fees, risk warnings.

The advisors would conduct the planning meeting, make recommendations, then spend [PLACEHOLDER] hours translating that into an SOA. Writing the narrative sections. Entering client data. Calculating fee impacts. Pulling in compliance language. Formatting. Proofreading.

On a client base of [PLACEHOLDER] clients, assuming [PLACEHOLDER] meetings per client per year (some more, some less), that was [PLACEHOLDER] SOAs annually. At [PLACEHOLDER] hours per SOA, that was [PLACEHOLDER] hours of advisor time—roughly [PLACEHOLDER] hours per advisor annually.

For advisors earning $[PLACEHOLDER] per hour (weighted between their billable rate and what you could charge clients for that work), that's roughly $[PLACEHOLDER] per advisor in opportunity cost.

But the real cost was in capacity. If an advisor spends [PLACEHOLDER] hours per client producing the SOA, they can only see so many clients. Each advisor was managing [PLACEHOLDER] clients. The natural question: How many clients could they manage if SOA production wasn't consuming [PLACEHOLDER]% of their time?

"I'm not a writer," one advisor said. "I'm good at talking to clients about their money. I'm not good at—and don't want to spend time—writing formal documents. But that's where all the time goes."

The practice was profitable. But it was constrained by manual capacity—the bottleneck was SOA production.

The Discovery

We examined the SOA production workflow in detail.

Client Data Entry: Extracting information from the meeting notes and client systems into the SOA template. Name, date of birth, contact details, current assets, current liabilities, income, goals, risk profile.

This information already existed—in the CRM system, in the client meeting notes, in the fact-find form. But someone had to manually transcribe it into the SOA template.

[PLACEHOLDER] minutes per SOA just to populate client data.

Narrative Writing: The personal advice narrative—an explanation of the advisor's analysis, why they'd made specific recommendations, how those recommendations addressed the client's stated goals.

This was the part that required the advisor's thinking. But it also required actual writing. [PLACEHOLDER] minutes per SOA for the advisor to compose a thoughtful narrative—and that was if they were efficient.

Recommendation Structuring: Translating recommendations into the formal SOA format—specific products, quantities, timing, expected outcomes.

[PLACEHOLDER] minutes, largely data entry and formatting.

Fee Calculation: Working through the fee impacts—what fees would the recommendations attract, how would fees be charged, what was the total cost to the client, what were the benefits relative to the fees.

[PLACEHOLDER] minutes of financial modeling and calculation work—some of it repetitive calculation that should have been automated.

Compliance & Risk Language: Inserting appropriate risk warnings, regulatory language, compliance statements. Most of this was boilerplate that applied to the recommendation type.

[PLACEHOLDER] minutes of copying and pasting standard language, with minor customization.

Document Assembly & Formatting: Taking all the sections—client data, narrative, recommendations, fees, compliance language—and assembling them into a professional-looking document with consistent formatting, page breaks, headers.

[PLACEHOLDER] minutes, mostly manual formatting work.

Proofreading & Review: Reading through the document to ensure accuracy, consistency, completeness.

[PLACEHOLDER] minutes, though often advisors would have to send it back to support staff for corrections, creating back-and-forth iterations.

When we looked at the [PLACEHOLDER] hours total, the breakdown was roughly:

  • [PLACEHOLDER] minutes of work that required advisor judgment (the narrative, the recommendation strategy)
  • [PLACEHOLDER] minutes of data entry, formatting, and boilerplate insertion that could be automated or templated

In other words: [PLACEHOLDER]% of the time was advisor expertise. [PLACEHOLDER]% was admin work that happened to be done by experts.

The Solution

We built an AI-assisted SOA system that handled the admin work and gave advisors the structure they needed to focus on the expertise part.

Data Auto-population: The system integrated with the practice's CRM and fact-find forms. When an advisor selected a client, the system automatically pulled client data—name, date of birth, contact details, current assets, liabilities, income, goals—and populated the SOA template.

Client data entry time: [PLACEHOLDER] minutes (system), down from [PLACEHOLDER] minutes (manual).

The advisor reviewed the data for accuracy but didn't have to manually type it.

Narrative Framework & Suggestions: This was the key innovation. The system didn't write the narrative. Instead, it created a structured outline based on the client's circumstances and the advisor's recommendations.

The outline looked like:

  • Client goals: [system-extracted goals]
  • Client circumstances: [system-summarized financial position]
  • Key considerations: [system-identified constraints or opportunities based on client data]
  • Recommended strategy: [system-templated recommendation framework]
  • Fee impact: [system-calculated fee summary]
  • Next steps: [system-suggested follow-up actions]

The advisor reviewed this outline, edited it to reflect their actual analysis and thinking, and within a structured framework, the narrative came together much faster.

The advisor no longer had to start from a blank page. They had structure. Narrative writing time dropped from [PLACEHOLDER] minutes to [PLACEHOLDER] minutes—the advisor was editing and customizing rather than creating from scratch.

Recommendation Auto-formatting: Once recommendations were decided, the system formatted them according to SOA standards—specific products, quantities, rationale, fee impact, implementation timeline.

The advisor typed the recommendation once; the system formatted it for the SOA.

Recommendation structuring time: [PLACEHOLDER] minutes, down from [PLACEHOLDER] minutes.

Fee Impact Calculation: The system calculated fee impacts automatically—based on the recommendations, the client's starting assets, assumed growth rates, and the practice's fee schedule, what would the fees be over different time horizons?

The advisor reviewed the outputs but didn't have to manually calculate.

Fee calculation time: [PLACEHOLDER] minutes, down from [PLACEHOLDER] minutes.

Compliance Language Auto-insertion: Risk warnings, regulatory language, compliance statements—these were templated and auto-inserted based on the recommendation type. The advisor saw them and could customize if necessary, but the boilerplate was already there.

Compliance language insertion time: [PLACEHOLDER] minutes, down from [PLACEHOLDER] minutes (now mostly just review).

Document Assembly & Auto-formatting: The system assembled all sections into a professional document with consistent formatting, page breaks, numbered sections, table of contents.

Formatting time: [PLACEHOLDER] minutes (the advisor just hit "generate PDF"), down from [PLACEHOLDER] minutes of manual formatting.

Built-in Compliance Checking: The system verified that all required elements were present—client declaration section, appropriateness statement, fee disclosure, risk profile, recommended action timeline. It flagged any missing elements before the document went to the advisor.

This reduced proofreading time and caught errors before they became document versions.

Total time per SOA dropped from [PLACEHOLDER] hours to [PLACEHOLDER] hours. The [PLACEHOLDER] hours of advisor expertise was still there. The [PLACEHOLDER] hours of admin work was gone.

The Results

SOA production time dropped from 12 hours to 3 hours per document. A 75% reduction.

That freed [PLACEHOLDER] hours per advisor annually—time that was previously consumed by document production.

With that freed capacity, each advisor took on additional clients. The practice went from managing [PLACEHOLDER] clients total ([PLACEHOLDER] per advisor) to [PLACEHOLDER] clients ([PLACEHOLDER] per advisor).

Assets under management grew from $[PLACEHOLDER] to $[PLACEHOLDER]—a [PLACEHOLDER] 31%]] increase in AUM with the same team and no added overhead.

But the real win was different: The advisors were now advisors instead of document writers.

One advisor put it simply: "I got back to what I'm good at. And the SOA is actually better—when I have time to think about it instead of time pressure to finish it, my recommendations are more thoughtful."

What's Next

The practice is exploring whether the same approach could apply to other compliance documents—financial situation reports, fund fact sheets, performance updates. The same pattern exists: structure + templates + data automation = time freed for actual thinking.

They're also testing whether the SOA system could generate a draft recommendation for the advisor to review and customize—not making the recommendation on behalf of the advisor, but offering a recommended strategy based on client circumstances, which the advisor could use as a starting point.

"The goal is always the same," a partner said. "We want advisors advising, not administrators administrating. That's when the advice is better and the business works better."

The Lesson

If you're running a professional services practice where your experts are spending [PLACEHOLDER]% of their time producing compliance documents, creating reports, assembling deliverables—work that requires their credentials but not their judgment—you have an AI opportunity.

The documents are necessary. The expertise matters. But [PLACEHOLDER]% of the time is probably admin work—data entry, formatting, boilerplate insertion, document assembly—that shouldn't require expert labor.

AI won't make your advisors or accountants or lawyers obsolete. But it will let them stop spending [PLACEHOLDER]% of their time on work that a system should be doing, and start spending [PLACEHOLDER]% of their time on work that only a human can do.

That's when capacity grows. That's when client service improves. That's when practitioners want to come to work instead of dreading document production.

Ready to find the document production and admin work that's keeping your experts off-focus? Book a free discovery call with CORSZA. We'll help you see it clearly—and then we'll give your experts their time back.