Microsoft Copilot gives registered investment advisors and financial services firms a way to move faster on analysis, compliance reporting, and client communication – without adding any staff. Firms managing SEC and FINRA obligations, quarterly client reporting cycles, and constant market monitoring have more to gain from this technology than most. This post covers practical use cases across Microsoft 365 tools, the security controls that matter for regulated firms, and what a responsible deployment looks like.
Key Takeaways
- Microsoft Copilot integrates natively into Word, Excel, Outlook, and Teams.
- Copilot only surfaces data the user already has permission to access, which is a foundational requirement for regulated environments.
- Microsoft 365 Business Premium includes data loss prevention policies, sensitivity labels, and legal hold capabilities that compliance teams require.
- 75% of Copilot users report being more productive, and organizations project a 112% ROI over three years.
Table of Contents
How Does Microsoft Copilot Support Financial Analysis?
Excel is where most financial analysis happens – from portfolio modeling to budget forecasting. Fortunately, Copilot works directly inside Excel. It does so in three distinct ways, each suited to a different type of task.
Reid Johnston, Teal Chief Intelligent Transformation Officer, demonstrated all three during recent Teal Microsoft Copilot virtual events. He was joined by Jill Williams, Teal VP of Finance and Accounting, who shared how she uses Copilot in her own day-to-day financial work. Let’s take a look at how you can use Copilot to save time.
The side panel: Querying your data, building formulas, and reconciling datasets.
The Copilot side panel in Excel handles three types of tasks that finance teams run into constantly.
1. Querying your data.
The first task we’ll cover is querying data in plain language. Analysts can ask questions like “Which holdings have the highest volatility relative to their benchmark?” or “Summarize the revenue trend across these business units” without writing formulas. Copilot reads the spreadsheet and returns a plain-English answer or summary.
2. Building formulas.
Jill uses formula generation in Copilot every week. In her workflow, she pulls trial balance data into a balance sheet by quarter, then asks Copilot to write the formula that matches account names across tabs and pulls the data over.
“I use Copilot a lot when I’m working on trial balances, working on balance sheets by quarter, breaking things down, and manipulating the data,” said Jill. “Sometimes I don’t know exactly the formula I want to use, because Excel has a ton of different formulas.”
Before Copilot, the process was exceptionally manual. She had to take multiple actions for each row of data. Now, Copilot simply generates the full formula, explains what it does, and offers alternatives if the first one does not fit.
Gar Whaley, Teal Chief Revenue Officer, noted that Jill is an expert Excel user – which means this is not a shortcut for beginners. This is a time saver for people who know their way around Excel.
3. Reconciling datasets.
When Jill needs to compare two datasets (e.g., two weeks of scorecard data or two periods of reporting), she prompts Copilot to reconcile the tables and surface what is different.
“I don’t want to go through the whole hullabaloo of looking at every single line,” said Jill. And she doesn’t have to.
Copilot flags mismatches and missing items, often generating output in a new sheet. The prompt can reference named tables for a cleaner result.
The =COPILOT formula: Putting AI output directly into cells.
This feature lets users type a prompt directly inside a cell to gain insight – just like any other Excel formula. So, you’d type:
The output populates the cell. Drag the formula down and Copilot runs the same prompt against every row in the column.
Reid demonstrated this with a customer CSAT spreadsheet. He typed:
After dragging the formula down, every response row had a sentiment label – such as Happy, Neutral, or Sad – generated in seconds. He then ran a second column, which generated recommended actions they should take for each sentiment label per row. Actions like, “thank the customer,” “conduct a customer service training session,” and “reach out with specific details.”
For financial services, the same pattern applies. You might tag client responses by sentiment, flag accounts by risk tier, or categorize transactions by type – all directly within the cells where the data lives, without exporting anything to a separate tool.
Agent Mode: Letting Copilot make changes to your workbook.
Agent Mode goes beyond querying. It lets Copilot actually modify the spreadsheet. For formatting tasks, you can type an instruction like, “Highlight every number greater than 10,000,” and Copilot applies the formatting across the entire dataset without conditional formatting menus or formula logic.
In the video, Gar Whaley raised a real use case for data cleanup. What if, after exporting credit card or financial transaction data, every number comes in as a negative? The fix is Agent Mode. You just type in the instruction, and Copilot rewrites the values.
To create a dashboard, Reid typed a single prompt asking Copilot to show trends and key points from the data. Copilot generated annual sales by business unit, expense trends by year, a profitability analysis, total revenue and profit KPIs, and visualizations – on a new sheet, without any manual chart setup.
Financial services firms might use this to create portfolio dashboards, budget vs. actual summaries, quarterly performance visualizations, or any task that typically involves an analyst spending an hour in Excel.
Specific capabilities in financial analysis include:
- Generating financial forecasts by referencing existing spreadsheet data and modeling assumptions without manual formula construction.
- Summarizing market trend data from documents or reports pasted into the prompt, then surfacing investment opportunities relevant to a client profile.
- Identifying anomalies in portfolio data (Copilot can flag positions that deviate from allocation targets and explain the deviation in plain English).
- Drafting narrative commentary for performance reports in Word, pulling figures directly from linked Excel data.
The efficiency gain is not about replacing an advisor’s judgment. In fact, quite the opposite. It is about removing the data wrangling that sits between raw numbers and an analysis that is ready for a client.
How Does Copilot Improve Client Advisory Services?
In Teams, Copilot can transcribe advisor-client meetings, generate a summary of what was discussed, and produce a list of action items and follow-up commitments. This allows advisors to spend meetings focused on the client, instead of documentation.
It can help you by:
- Personalizing investment recommendations using Copilot in Word to draft client-specific financial planning documents that reference the client’s stated goals, risk tolerance, and current portfolio.
- Generating client education content. Copilot can draft explainers on topics like tax-loss harvesting, bond duration risk, or Roth conversion strategies at a reading level appropriate for a non-expert audience.
- Drafting quarterly or annual client update emails in Outlook that reference recent market events and connect them to the client’s portfolio positioning, rather than sending a generic firm-wide letter.
- Summarizing meeting notes in Teams so CRM systems can be updated quickly after each client interaction, reducing the administrative backlog that builds up over a week of meetings.
According to Microsoft WorkLab data, 79% of small and midsize business executives believe AI is important to staying competitive. For financial services firms competing with larger RIAs and wirehouses that have had AI tools for years, Copilot is a great way to close that gap.
Is Microsoft Copilot Secure Enough for Financial Services?
Security is the first question any compliance-minded firm should ask before deploying an AI tool. Copilot in Microsoft 365 is built on the Microsoft commercial cloud, which means several specific protections apply by default, but there is one step firms need to take before turning on licenses.
Review data access before you flip the switch.
Reid Johnston raised this directly when asked what firms should do before implementing Copilot.
“Before you turn on the licenses, make sure that you know who has access to your data,” said Reid. “Because Copilot will surface that information that your users may not know they already have access to. If a junior analyst has inadvertent access to a client folder they were never supposed to see, Copilot will surface that content when they ask a related question.”
So, fixing permissions before enabling Copilot is the first step. Reid also offered some broader guidance.
“Before you implement any AI solution, get together as an organization and identify what the rules are going to be,” he said. “Set the rules first.”
Gar explained that Teal’s framework for getting clients set up with AI includes: AI readiness, policy, and governance – in that order – before deployment begins.
What the security architecture covers.
This architecture addresses the foundational concerns compliance officers raise most frequently. It is not a substitute for a firm-specific security review, but it establishes the right starting point.
- Copilot only accesses the data that the user already has permission to see.
- Your organization’s data does not train Microsoft’s large language model.
- Microsoft 365 Business Premium includes data loss prevention policies that prevent sensitive client data from being included in Copilot prompts or exported in ways that violate firm policy.
- Sensitivity labels applied to documents flow through Copilot interactions. If a document is labeled Confidential, that classification is respected in how Copilot handles the content.
- Legal hold and eDiscovery capabilities in Microsoft 365 cover Copilot interactions, supporting recordkeeping obligations under financial services regulations.
What ROI Should Financial Services Firms Expect?
A Forrester Total Economic Impact study (commissioned by Microsoft) projected a 112% return on investment over three years for organizations that deploy Copilot. That figure reflects time savings across writing, summarizing, and data analysis tasks – which are the tasks that consume the most billable and non-billable hours in a financial services firm.
The 75% productivity improvement figure in the study is self-reported, and individual results will vary based on how thoroughly staff are trained and how well the tool is integrated into existing workflows. That means firms that treat Copilot as a drop-in tool without a structured adoption plan consistently see lower returns than firms that invest in onboarding.
Deploying Microsoft Copilot with the Right Controls
Microsoft Copilot works best for financial services firms when it is deployed on a foundation that includes the right licensing, security configuration, and staff training. Microsoft 365 Business Premium provides the compliance controls that regulated firms need before turning Copilot on.
Teal has the experience supporting regulated firms with managed IT services for over 25 years, and can help you adopt AI safely and efficiently. We hold ISO 27001 certification and help firms deploy Microsoft Copilot with the security configuration you need. Contact us anytime to start a conversation about licensing and data governance.
