Restaurant Performance Improvement Plan

Capability, not conduct

A restaurant performance improvement plan is a time-boxed written plan that names a specific performance gap, sets measurable targets, commits the restaurant to specific support, schedules check-ins on real dates, and states the consequence if the targets are not met.

The distinction that matters: a PIP is for an employee who is trying and falling short. Discipline is for an employee who chose the behavior. A cook who cannot hold ticket times yet needs a plan. A cook who walked out Friday needs a write-up. Running them backwards is how restaurants end up with final warnings for slowness and coaching notes for no-shows.

Build the plan with the PIP generator — it computes the check-in dates from your start date so they exist before anyone signs.

Decide what this document is for

Before writing anything, answer honestly: if this employee hit every target, would you keep them?

If yes, you are writing a development plan. Write it to be passed — realistic targets, real training, a trainer with a name.

If no, a PIP is the wrong instrument. You are documenting a decision, and thirty days of pretending otherwise costs the employee time and costs you credibility with the rest of the team, who will read the outcome correctly. Handle it as a separation with the documentation that supports it, or fix the actual problem — which is sometimes the schedule, the station, or the training the restaurant never delivered.

Targets from data you already have

A restaurant is unusually well instrumented. Use it.

Role	Measurable target	Source
Line cook	Average cook time on sauté under 9 minutes on Fri/Sat dinner	Kitchen display or ticket times
Server	Void percentage at or under 2.5%; zero unapproved comps	POS void/comp report
Server	Dessert or wine attach rate at or above 18%	POS item mix
Bartender	Cash variance within $5 per shift for 10 consecutive shifts	Drawer counts
Prep	Spec accuracy on 5 audited recipes with zero yield variance beyond 5%	Prep sheets and yield log
Host	Quote accuracy within 5 minutes on 90% of quotes	Reservation/waitlist system
Shift lead	Side work checklist completed and verified on 100% of closes	Closing checklist
Any	Zero guest complaints naming the employee over the plan window	Complaint log

Each target needs three things: a number, a source you will actually pull, and a date. If you cannot say where the number comes from, the employee cannot self-monitor, and the check-in becomes an argument about impressions.

Two to four targets. A plan with nine targets is a plan the employee cannot prioritize and you will not measure.

What the restaurant commits to

This section is what separates a real plan from a formality, and it is the part reviewers read closely if the plan ends in separation.

• Training shifts — how many, on which station, and with whom by name.
• Materials — specs, steps of service, the pour guide, the allergen matrix. Given, not "available."
• Scheduling — enough shifts on the station being measured. You cannot fairly measure sauté speed while scheduling someone on garde manger.
• Feedback cadence — the check-ins, on the dates below.
• Who to ask — the specific person the employee goes to mid-shift when they are stuck.

If the plan window closes and the restaurant delivered two of five committed training shifts, the plan did not test what it claimed to test. Note that honestly rather than proceeding as if it did.

Check-ins on real dates

Set them when the plan is signed, as calendar dates.

Plan	Check-ins
30 days	Day 10, day 20, day 30
60 days	Day 15, day 30, day 45, day 60
90 days	Day 30, day 60, day 90

Each check-in is short and written: the target, where the number actually landed, what changed since the last one, what support is still owed, and whether the plan continues unchanged. Five sentences is fine. Nothing is not.

A missed check-in is worth logging too — "day 20 check-in moved to day 23 due to scheduling" is a fact. Silently skipping it and then citing the plan at the end is what makes an employee feel ambushed, and it is visible in the record.

The closing conversation

At the end of the window, one of three things is true. Say which, in writing.

Targets met. Close the plan explicitly. Put the closure in the record — an employee who improved should not have an open-ended PIP following them around. Note what changed and what standard now applies.

Partial progress. Extend once, with revised targets and new dates, and say why the extension is warranted. Do not extend indefinitely; a third extension means the targets were wrong.

Targets not met. Reference the plan, each check-in, the actual numbers, and the support delivered. That record is what makes the next step defensible — and it is far stronger than a file with a single "not a good fit" memo.

Where PIPs fail in restaurants

• Targets written as adjectives: "be faster," "show more urgency," "improve attitude"
• Nine targets nobody can prioritize
• Check-in dates left as "weekly" and never held
• Training committed on paper and never scheduled
• Employee scheduled off the station being measured
• Used for conduct — no-shows, refusals, policy violations — instead of discipline
• Written after the decision was made, which the whole staff figures out
• No closure note when the employee succeeds
• The plan lives in the GM's laptop and disappears when the GM transfers

Conclusion

A PIP is a test the restaurant has to help the employee pass. Two to four measurable targets from data you already pull, the support you commit to in writing, check-ins on real calendar dates, and an explicit close at the end either way.

The PIP generator computes the check-in dates and holds the target rows. Before a PIP, most performance gaps deserve a documented conversation first: restaurant employee coaching. For the documentation standard behind all of it, the pillar guide.

Docu-Coach™ keeps the plan, the check-ins, and the coaching that preceded it in one employee history — better documentation and more consistent records, not legal advice. See it at /demo.