Before you touch a missing value, decide what it means. A missing city is unknown. A missing discount at Juniper means no discount was given. The two need different handling.
replace_na() fills gaps with a value: mutate(city = replace_na(city, "Unknown")), or for several columns at once, replace_na(list(city = "Unknown", segment = "retail")). drop_na(city) drops the rows where that column is missing; drop_na() with no columns drops every row with any gap, which usually throws away far more than you meant.
coalesce(a, b) takes the first value that isn't missing, row by row: coalesce(mobile, landline). It is also a short way to write "replace missing with 0": coalesce(discount, 0).
Fill a gap with a label, drop rows missing one column, or fall back to a value.