The person who posted this question has the right instinct but the wrong tool. They are trying to solve a data synchronization problem with a spreadsheet, and spreadsheets were never built for that job. What they describe, linking a master catalogue of home-building materials across tens of spreadsheets, then swapping placeholder SKUs for vendor-specific codes, is not a spreadsheet task. It is a database task wearing a spreadsheet's clothes.
Let's be clear about what is happening here. The user has a set of templates that use generic "in-house" SKUs because those templates serve multiple vendors. Now one vendor has provided a real catalogue, and the user needs every instance of SKU 1234 (drywall) to become SKU 4567 across dozens of sheets. In a traditional spreadsheet, this means manual find-and-replace, fragile VLOOKUPs, or macros that break the moment someone sorts a column differently. None of those scale. None of those survive contact with a second vendor. The user is essentially asking how to turn a static document into a living data network, and the answer is not a better formula.
What this person actually needs is a system that treats the SKU mapping as a single source of truth. Imagine a master catalogue that lives in one place, and every template simply references it. When the vendor updates a SKU, every sheet that draws from that catalogue updates automatically. No manual propagation. No error-prone copying. That is the model an AI-native spreadsheet enables: data that flows instead of data that sits. The user's current workflow treats SKUs as text to be replaced; the better workflow treats SKUs as relationships to be maintained.
The practical takeaway is this: if you are managing multiple spreadsheets that all depend on a common reference list, stop treating each sheet as an island. Look for a tool that lets you define that reference once and then link every template to it. The technology exists now. The barrier is not capability, it is the habit of opening Excel first. The user who posted this question is already halfway there. They recognized the problem. The next step is choosing a solution that matches the complexity of the task, not the comfort of the old interface.