Copilot in Excel Will Soon Wrangle Your Data with Plain English

If you have ever spent a Monday morning copying figures out of a CSV, fixing dates that refuse to format properly, and hunting down duplicate rows before a report is due, this one is for you. Microsoft has added a new item to the Microsoft 365 roadmap that promises to take a lot of that grunt work off your plate: Copilot in Excel will soon be able to connect to external data, transform it, and load it into your workbook, all from a plain English request.

What is changing?

Copilot in Excel turning a messy stack of files into a clean, refreshable table

Today, getting external data into Excel usually means one of two things. Either you copy and paste and then spend ages tidying up, or you venture into Power Query, which is brilliant once you know it but intimidating if you do not. The roadmap update describes a third option: you simply tell Copilot what data you need and how you want it prepared, and it does the connecting, cleaning and loading for you.

The important detail is how it works behind the scenes. Copilot does not just paste static numbers into your sheet. It builds a proper Power Query underneath, the same kind you would create yourself. That means the result is refreshable, auditable and editable. When next month's file arrives, you refresh the query rather than starting again, and if something looks odd you can open the query and inspect every step.

What this looks like in practice

A few scenarios where this could genuinely save time:

  • The HR coordinator's staff list. Every month, HR exports a CSV from the payroll system with inconsistent capitalisation, dates in three formats and the occasional duplicate. Instead of a Friday afternoon of find and replace, they ask Copilot to pull in the file, standardise the date columns, proper-case the names and remove duplicates. Done, and refreshable next month.
  • The project manager's status report. Project data lives in a SharePoint list, but the steering pack is in Excel. Rather than exporting and reformatting weekly, they ask Copilot to connect to the list, keep only active projects and group them by workstream. The underlying query refreshes whenever the list changes.
  • The finance analyst's monthly actuals. Actuals arrive as a workbook from another system with extra columns and a header row that never sits still. Copilot can be asked to import the sheet, skip the noise, rename the columns and append the result to the year-to-date table.
  • The operations lead combining regions. Three regional workbooks, three slightly different layouts. A plain English request to stack them into one tidy table beats an afternoon of manual surgery.

Why the Power Query part matters

It is worth dwelling on this, because it is the difference between a party trick and a tool you can rely on. AI-generated answers can be hard to trust when you cannot see how they were produced. Because Copilot builds a standard query, your team keeps full visibility: open the Power Query editor, review the applied steps, adjust anything that is not quite right, and the logic stays with the workbook. It also means the skills your team already has, or is building, remain relevant. Copilot gets you started faster; it does not replace understanding what the query is doing.

Who does this affect?

Anyone who regularly imports and cleans data in Excel, which is most of us at some point. Business users who have avoided Power Query because the interface felt like deep water get the most immediate benefit. Experienced analysts get a faster starting point for repetitive wrangling. And IT teams should take an interest too, because connecting to external data still requires the right permissions, which brings us to preparation.

How to prepare

The roadmap lists this as planned, with availability targeted around October 2026. As always with roadmap dates, treat that as an intention rather than a promise, since timelines do slip. In the meantime, a few sensible steps:

  • Build baseline Power Query skills. Copilot produces standard queries, so someone on the team should be comfortable reviewing applied steps, checking data types and fixing refresh errors. A little knowledge goes a long way here.
  • Review data source permissions. Copilot can only reach what the user is allowed to reach. If your SharePoint lists, shared drives or line-of-business exports have messy permissions, tidy them before people start connecting Copilot to them.
  • Audit your manual reports. Make a list of the regular reports that currently rely on copy, paste and prayer. Those are your prime candidates, and having the list ready means you can test the feature properly the moment it arrives.
  • Set expectations. Let people know this is coming and what it is good for, so the first use is a planned one rather than someone pointing it at something sensitive.

The bottom line

This is one of the more practical Copilot additions we have seen on the roadmap, because it targets a chore almost everyone recognises. The fact that it produces real, inspectable Power Queries rather than opaque output makes it much easier to trust and to govern. We will keep an eye on the rollout and share hands-on impressions once it lands.

If you would like to build practical skills across Excel, Copilot and the wider Power Platform, explore 3grow's hands-on training courses at https://3grow.net/courses/.