I think similar to “code smells” in software engineering, using excel for critical data pipeline work can be considered a “data smell”. It’s not bad in and of itself, but it’s a signal that something may have gone seriously wrong with the data engineering process.