A coloured dropdown looks like a controlled list, but the rule can either reject an unlisted value or merely warn about it. Choose that behaviour for the task. A strict status column and a column collecting new suggestions have different needs.
Build a small example first
In a test sheet on a computer, select A2:A5 and open Data, Data validation, Add rule. Create a dropdown with two entries, New and Done. Keep it single-selection for this example. Use an empty range so you can observe the rule without changing existing records.
In Advanced options, inspect the treatment of invalid data. The default dropdown rejects an item outside the list. Show a warning allows that entry while flagging it. Menu names depend on the interface language; mobile controls differ.
Check an allowed value and an exception
Select New in A2. In A3, try typing Later, which is not on the list. With rejection active, confirm it is not accepted as an ordinary valid entry. On the test copy, switch to warning and try again: check that the value remains with a warning. Then restore the behaviour you actually want.
Keep the rule separate from data quality
For a working status column, keep a short agreed vocabulary and explain the exceptions. For suggestions, a warning may be useful while new terms are reviewed. Neither the chip colour nor the presence of a dropdown proves that older data are correct. Check existing values separately before using the column in a count or report.