This is exactly the kind of problem that should have been solved years ago, and the fact that it still requires a Reddit post and a custom formula tells you everything about how far behind traditional spreadsheets have fallen. AppleKat89 has two systems that speak different languages, one stores item numbers as clean digits, the other buries them inside a code like Q12345XVTE, and the only bridge between them is manual effort or a fragile workaround. That is not a user error. That is a tool failure.
What AppleKat89 needs is not a clever string of nested functions. She needs a system that understands data the way people do: by meaning, not by format. When a supplier statement says "12345" and her inventory system says "Q12345XVTE," any human can see the connection instantly. The spreadsheet should be able to do the same. The practical ask here is simple: extract the numeric core from a mixed string, match it against a clean number, and return the corresponding quantity. Excel can do this with a combination of MID, FIND, and INDEX-MATCH, but that solution is brittle. It breaks when the prefix length changes, when the suffix disappears, or when a different supplier uses a different pattern. The user ends up maintaining the formula instead of using the result.
This is where an AI-native approach changes the game. Instead of forcing the user to reverse-engineer every pattern, the tool can learn the logic of the match. It can recognize that "12345" is the common thread across both records, ignore the surrounding text, and automate the reconciliation. The outcome is not a faster formula, it is the elimination of the formula entirely. The user asks, "Do these quantities match?" and the system answers. That is the difference between managing data and letting data work for you.
The real cost of the old approach is not the time spent writing the formula. It is the time spent testing it, debugging it when a new supplier arrives, and re-explaining it to the next colleague who inherits the file. AppleKat89's question is reasonable, but the fact that it has to be asked at all is a sign that the tools we rely on have not kept pace with the work we actually do. The solution is not a better workaround. It is a spreadsheet that treats data extraction and matching as native capabilities, not as puzzles for the user to solve. That is the standard we should expect, and it is the one we are building toward.