When the same step applies to several columns, across() saves writing it out each time. It takes the columns, then a function: summarise(across(c(hours, hourly_rate), mean)).
Pick columns by name, by a range (q1:q4), with helpers such as starts_with("q"), or by type with where(is.numeric).
To pass an argument, write a short function: across(c(minutes, distance_km), \(x) mean(x, na.rm = TRUE)). \(x) is R's shorthand for function(x).
With several functions, give them names in a list and choose the column names with .names: across(c(revenue, customers), list(total = sum, avg = mean), .names = "{.fn}_{.col}") makes total_revenue, avg_revenue and so on. Inside mutate(), across() changes the columns in place unless .names gives new ones.
The same summary over several columns, with an argument passed through.