Anyone who spends time in Excel knows the real challenge isn’t usually the calculations. It’s getting information into the format you need.
Reports often need specific headings. Financial figures may need to be displayed in a particular way. Teams have their own naming conventions, formulas, layouts and processes that everyone is expected to follow. Individually, these tasks don’t take very long. Collectively, they consume a surprising amount of time.
And that’s exactly the kind of repetitive work AI is supposed to help with.
Microsoft has been steadily expanding Copilot’s capabilities within Excel, and some updates have been more useful than others. This latest development feels particularly practical because it focuses on something many users do every day: repeating the same instructions.
The update allows users to teach Copilot their preferred way of working. Whether that’s how currency should be displayed, how reports should be structured, or how specific types of data should be formatted, Copilot can learn those preferences and apply them automatically in future.
For anyone who’s found themselves making the same formatting changes over and over again, that could be a significant time saver.
There’s another element of the update that could have an even bigger impact for teams.
Most businesses rely on shared spreadsheets, and consistency can often be a challenge. One person creates a report one way, someone else structures it differently, and before long there are multiple versions of what should be the same document.
Microsoft is introducing workbook rules that stay with the spreadsheet itself. These rules help Copilot understand how a particular workbook should be used and formatted. If the file is shared with colleagues, those instructions travel with it.
That’s important because it reduces the need for constant explanations and helps create consistency across teams. Instead of relying on everyone remembering the process, the workbook itself provides guidance to both users and the AI assistant.
What’s particularly interesting is what this says about the direction of AI.
Until now, most people have used tools like Copilot to answer questions, generate content, explain formulas or summarise information. Useful? Absolutely. But this feels like a step beyond that.
Rather than simply responding to prompts, the software is starting to learn how a business prefers to work. The more context it has about processes, formats and preferences, the more relevant and useful the output becomes.
Of course, that doesn’t remove the need for human oversight.
Spreadsheets often contain critical business information, and any AI-generated output should still be reviewed before it’s relied upon. That’s especially true when financial reporting, formulas or business-critical data are involved. AI can help with the work, but responsibility for accuracy still sits with the people using it.
That said, for the everyday tasks that quietly eat away at productivity, this feels like a sensible step forward.
If Excel can spend less time asking how something should be formatted and more time helping people get meaningful work done, that’s the sort of AI adoption most businesses can get behind.