Recommended
Excelerate Summit Homepage (Pre-Event)
Gain exclusive insights into cutting-edge AI technologies, connect with visionary leaders in strategic finance, and explore the next-gen tools reshaping the future of finance.
As a finance team, you're always working against the clock.
While closing on day 10 might have worked 10-15 years ago, today, you're expected to get it done by day three or even earlier to support the business's forecasting needs.
On top of these shifting expectations, there's what exists outside of the job description--what sounds like a simple request from an executive leader to find room in the budget can turn into several days of work for your FP&A team.
With recent advances in generative AI such as ChatGPT, however, finance professionals may be able to get this kind of work off their plates. And that's exactly what Microsoft is proposing to do with its new Copilot feature.
Microsoft announced Copilot on March 16, generating a flurry of excitement as the AI assistant promises some of the most meaningful enhancements to productivity since the introduction of the web and cloud.
Essentially, Microsoft is combining the power of generative AI with the tools most organizations already use today—its suite of Microsoft 365 apps.
Copilot integrates directly with apps, such as Word, Excel, PowerPoint, Outlook and Teams, to help remove the drudgery of work. With natural language processing, you can instruct Copilot to summarize your meeting, draft a document, analyze data in a spreadsheet, draft an email and more. The idea is to "take away the shallow work so that humans can do the deep work that we really crave," according to Microsoft's Vice President of People Strategy Nate Boaz.
Finance departments are no strangers to drudgery, with a significant amount of their time taken up by manual work. In our 2022 benchmark report, The State of Strategic Finance, 73% of the businesses we surveyed said they spend too much time on manual processes associated with planning and budgeting activities.
With most finance professionals being Excel power users, Microsoft Copilot presents an exciting outlook (no pun intended) for finance teams hoping to reclaim their time and play a larger role as strategic partners to the business.
Copilot is powered by large language models (LLMs), a type of machine learning trained to recognize, summarize and generate text. ChatGPT, created by OpenAI (which is backed by Microsoft), is one of the main building blocks of Copilot.
As Microsoft explains in their announcement, Copilot's transformational nature lies in its ability to be grounded in your organization's unique content and context. It does this through the Microsoft Graph, which is the secure environment where your business information (emails, meetings, chats and calendar) lives.
When you give Copilot a prompt in one of your Microsoft 365 apps (say, "Summarize this quarter's results in three key trends" in Excel), it enriches that ask with context from your business data in the Microsoft Graph. This helps improve the quality and relevance of the answer. Then, it'll send this modified prompt to the LLM, followed by the Microsoft Graph once again. Finally, Copilot will send the response back to you, along with commands for the app you're using.
All this happens behind the scenes in a matter of moments.
In Excel, Copilot can help you make sense of your data exceptionally fast.
This is especially game changing for FP&A professionals because to analyze your company's financial performance at a glance, you won't need to whip out a series of formulas. Instead, you can ask Copilot questions in natural language and it'll identify trends and correlations for you.
It also means you can answer questions from across the company faster, helping you serve as a better business partner.
Source: Microsoft
Here are four examples of how Copilot will bring value to finance teams:
CEOs depend on their finance leaders to provide updates on the business's performance at a moment's notice. This is much harder to do when you have no easy way to get a sense of what your data is telling you.
Rather than spending time sifting through your data set to find trends and correlations, you can simply ask Copilot, which can provide you with a starting point.
What's more, you can have Copilot create charts and tables that illustrate responses to your prompts in greater detail.
For instance, you can ask Copilot to give you a breakdown of your sales for the quarter by region, and it'll generate a table for you in a new sheet. This allows you to dig deeper into your data without actually modifying it.
Let's say you're about to head into a meeting and get a last-minute request to pull together some insights on how each of your product lines are performing quarter over quarter. With Copilot, you can pull together an ad hoc report in minutes with just a few prompts.
Copilot can also apply conditional formatting such as color coding to your data visualizations to make trends and insights easier to spot.
You can also save time on formatting reporting in PowerPoint with Copilot. Instead of spending time adjusting each slide of your presentation, you can use natural language commands to apply any formatting needs to the entire slide deck.
These features ultimately let you spend less time making your reports presentation ready and more time interrogating the data itself and forming insights.
You can even ask Copilot what-if scenarios based on the data in your Excel spreadsheet. In addition to giving you the results of a scenario, Copilot can create a financial model based on the variables you gave it and visualize it with a graph or table.
Copilot's answers will include an "Explain" button, so you can double-check its work by asking it to break down how it created a model.
Microsoft stresses that this next generation of AI with Copilot is about "augmentation" rather than "substitution." Although powerful, the likes of Copilot and ChatGPT can't replace what a human brings to the table.
The output of generative AI is also not perfect, which means there will always need to be a layer of human verification involved.
In a research blog about GPT-4 (which fuels Copilot's LLM), OpenAI states, "despite its capabilities, GPT-4 has similar limitations as earlier GPT models. Most importantly, it still is not fully reliable (it "hallucinates" facts and makes reasoning errors)."
By "grounding" prompts with additional business context from your Microsoft 365 apps, however, Copilot reduces the likelihood of these errors.
A popular concern among finance professionals regarding the recent AI boom is whether or not their role will be replaced. That's definitely not the case here and in fact, having an AI assistant like Copilot means your finance team will be able to focus on more value-adding work.
The time it takes your financial analyst to wade through large amounts of data and make sense of it leaves little time left for deep analysis. But with the ability to carry out complex Excel operations with natural language alone, your finance team can ascend beyond number crunching to serve as your company's strategic powerhouse.
What we often hear from the finance teams we talk to is that they're not able to fully participate in the business's strategy-forming sessions because they have too much manual work to process. But with tools like Copilot freeing up that thinking time in a meaningful way, this will change the role of Finance in the organization to that of a desired strategic partner.
Being able to manipulate your financial data more quickly can also lead to faster forecasting and more agile planning.
In our 2022 benchmark survey, we found that only 29% of the businesses we surveyed were forecasting on a monthly basis, with 34% only forecasting once or twice a year. This puts a damper on businesses' ability to make fast decisions, mitigate risk and take advantage of new opportunities.
The most commonly cited challenges getting in the way of those companies' ability to forecast more often were: manual and error-prone forecasting processes (cited by 48% of respondents), insufficient or ineffective technology (23%) and insufficient skill sets or experience on the planning team (19%). These are all areas where Copilot could provide a boost.
Another area where skill sets can be a gap within finance teams is data visualization. Regarding the technical skills finance teams need to thrive, our survey found that just 24% of respondents were currently prioritizing data visualization. Despite this, 49% of respondents said data visualization was a priority area for adding strategic value, signaling a need to close that skill gap.
Copilot in Excel presents the possibility of making data visualization much more accessible. Instead of waiting for your team to get over the learning curve of creating visualizations in Excel, you can have Copilot generate these visuals for you with natural language prompts.
Data visualizations are an important aspect of how finance departments communicate the meaning of their reporting to the rest of the company, making the advancements Copilot can bring here especially compelling.
We have a deep relationship with Microsoft, a core technology partner of Vena for over a decade. As Microsoft releases new capabilities, we work closely with them to make those features available to our customers.
We're also eagerly embracing the new capabilities of OpenAI, who is incidentally, a Vena customer.
Since Vena is natively integrated with Excel, Microsoft's announcement of Copilot—leveraging OpenAI's ChatGPT—naturally has us very excited at the possibilities for our customers.
With Vena allowing you to take your spreadsheets and templates and connect them to a data model in the cloud, adding Copilot into the mix has even more powerful implications for managing the lifecycle of FP&A processes.
Microsoft has not yet fully disclosed when Copilot will become available, but says they will be "expanding these previews to customers more broadly in the coming months." Microsoft is currently testing Copilot with 20 customers, including eight Fortune 500 companies.
