This user is asking the wrong question. The real issue isn't how to strip text from a cell, it is why anyone should have to build a workaround for a problem that should not exist in the first place.
The scenario is painfully familiar. A team builds an app for practical use. They include both customer name and number for clarity. Then they realize the downstream system only needs the number. So now they face thousands of rows like "1234567 Cable Coffee Shop" and need a formula to delete the words. The assumption is that Excel can handle this if you find the right function. And technically, it can. A combination of LEFT, FIND, and some array logic will isolate the digits. But let us be honest about what this means in practice: you are spending time writing and debugging a formula for a task that your tools forced upon you. The spreadsheet did not create the problem, but it also did nothing to prevent it.
This is where we think the conversation needs to shift. The user is not asking for a better formula. They are asking for a tool that understands the difference between a customer number and a customer name. They are asking for a system that can separate data at the point of entry, not at the point of cleanup. Traditional spreadsheets treat every cell as a container of raw text. They have no awareness that "1234567" and "Cable Coffee Shop" serve different purposes. So the burden falls on the user to write conditional logic, test edge cases, and hope the formula holds across thousands of rows. That is not productivity. That is maintenance disguised as work.
An AI-native approach changes the premise entirely. Instead of asking "how do I remove the name," you should be able to say "keep only the customer number." The tool should recognize the pattern, digits followed by text, and execute the transformation without a formula. More importantly, it should have prevented the mixed data from entering the workflow in the first place. When you design a feed and return process, the system should enforce structure, not flexibility. The user should not have to choose between readability for humans and usability for machines.
The solution this person needs is not a better Excel function. It is a tool that treats data with the same respect they do. Strip the name if you must. But the real win is building workflows where that step is never necessary.