Since the 20% really is a commission for sales, is that better or worse than if we went to a system where servers were just paid a salary or wage per hour, with pay level based on performance?
I get that there are definitely certain types of sales positions where commission is preferred because you can make a lot more money that way. But I'm wondering if restaurant service falls into that category.
The current tipping system is an attempt at performance based pay, is it not? Theoretically, the better service the server provides, the better tips s/he will earn. If we go to a straight 20% of sales for all servers, that eliminates the performance aspect of it.
From my perspective, if a server is giving poor service, the natural consequence would be to assign them to hours and sections of the restaurant where they will impact the fewest customers, not cut their pay. In a commission-based system like the OP restaurant, this would almost certainly have the side effect of lowering their pay as well. If someone is giving poor service, then lowering their pay doesn't do anything concrete to help the customers get better service. All it does is provide a nebulous motivation to "improve."* OTOH, Adjusting the shift assignments so customers are only stuck with the "bad" servers when there are no "good" servers available (i.e., during off hours, when the restaurant is really busy, etc.) does have a direct impact on service. And I would expect that being stuck with less desirable shifts (and the resulting reduced sales commissions) would provide motivation to improve just as well as a direct pay cut.
*Another issue with adjusting pay based on performance is that you would need a very objective measure of performance. Otherwise, an employee's wages would be at the whim of whoever assesses performance, and I think that would open the business up to some nasty legal battles. Sales is one possible measure, on the assumption that improved sales means the customers are getting prompt and efficient service that allows for high table turnover and/or happy enough to linger over additional courses/drinks, or agree to upsells suggested by the server. It's also a measure that is directly linked to the businesses profits. If you wanted to get away from the commission model, you'd need other objective performance measures, and I'm not sure what they would be.
I'm not saying their pay should be cut for bad performance, quite the opposite. In the performance-based pay model, good servers would get pay increases for good service. Bad servers would get coached on how they could improve and eventually get fired if their performance didn't improve.
At least, this is how it works at my job. I get periodic reviews and raises/bonuses if I'm doing well. People who aren't performing up to par get feedback on what needs improvement and eventually, if they don't improve, they are let go.
How objective is the measure of performance? Well, it's not. It's subjective. It's based on how my boss perceives my performance. I don't work in a situation where output can be measured, as in, how many widgets I produced and whether they met standards. Similarly for servers, it would be subjective -- based on the manager's observations on how well the server is meeting the customers' needs.
As far as "Adjusting the shift assignments so customers are only stuck with the "bad" servers when there are no "good" servers available (i.e., during off hours, when the restaurant is really busy, etc.)", aren't you in fact, punishing the customers who come during those off-hours, etc. by providing them with a server you know is "bad"? That doesn't seem like a good business practice to me.
Again, I'm not necessarily saying performance-based is the way to go vs. commission. I'm just wondering if commission is really the best compensation method for the restaurant model.