The Improvement Plan That Gives a Fair Last Chance

Why this matters

A performance improvement plan, or PIP, has a bad reputation because most are written as a formality on the way to a firing, everyone in the room knows it, and it fixes nothing. That is a waste of a genuinely useful tool. A real improvement plan is a fair, honest last chance: a written agreement that says exactly what has to change, exactly how you will help, exactly by when, and exactly what happens if it does not. Built that way it does two hard things at once. It gives a struggling person a clear and winnable path back, and it creates the record that protects everyone if the path is not taken. This is the anatomy of one worth writing.

First, make sure it is a fair chance, not a countdown

Before you write a word, decide honestly which one this is. If you have already made up your mind to exit the person and the plan is just paperwork to justify it, do not dress it up as a chance, because the crew and the tech will both see through it and you will have taught everyone that your process is theater. A fair plan means the outcome is genuinely open: if they hit the marks, they stay, no asterisk. Only build a plan you would be glad to see succeed. If you would not, you are not writing an improvement plan, you are documenting a decision, which is a different and legitimate thing, just not this.

Write goals specific enough to pass the clarity test

The heart of a good plan is targets so concrete that the tech and you would independently agree whether each was met. Vague goals ("improve quality," "better attitude") make the plan unwinnable and unfair, because success is left to your mood. Convert every goal into an observable, measured standard:

• Not "be more reliable," but "on site by start time every scheduled day, with a call to dispatch before start if delayed."
• Not "reduce callbacks," but "callback rate back inside the shop's healthy range across the plan window."
• Not "work faster," but "standard maintenance calls trending to the expected pace, with hard finds noted."

Keep it to a small number of goals, two or three, not ten. A plan with ten targets is a plan designed to fail, because nobody can move ten things at once.

Name the support you will provide, in writing

This is the half that separates a fair plan from a trap, and the half most owners skip. Improvement is a two-way obligation. If you demand a change without providing the means, the plan is rigged. Put your commitments in the document alongside theirs:

• The training, ride-alongs, or pairing with a strong tech you will set up.
• The tools, checklist, or reference materials they need and have not had.
• A regular check-in cadence so they get feedback during the window, not just a verdict at the end.

A plan that lists only the employee's obligations reads as a setup. A plan that lists yours too reads as a partnership with a hard deadline, which is what it should be.

Set a real timeline with checkpoints inside it

Pick a window long enough to show real change and short enough to force it, commonly on the order of a month or two for most performance issues, longer for a genuine skill build. The length is a judgment call, not a rule, so match it to how long the change realistically takes. Then put checkpoints inside the window, not just an ending. Weekly or biweekly touch-points do two things: they give the tech a chance to correct course before the final date, and they keep the final review from being a surprise. A plan with no interior checkpoints is an exam with no study period.

State the consequence plainly

The tech has to know what is at stake, or the plan carries no weight. Say it in plain language: what happens if the goals are met (the plan closes, they continue), and what happens if they are not (the specific next step, up to and including ending employment). Do not soften this into fog. The honesty is a kindness. A tech who does not understand that this is a real line will not take the plan seriously, and then the failure lands as a shock instead of a known risk they were given every chance to avoid.

Deliver it as a conversation, not a served document

How you hand it over sets the tone for the whole window. Sit down, walk through it together, and frame it as what it is: "I want this to work, here is exactly what it takes, and here is how I am going to help." Have the tech acknowledge it, ideally with a signature, and note "declined to sign" if they will not, since acknowledgment matters for the record. Give them a copy. A plan slid across a desk in silence signals the countdown; a plan walked through with genuine intent signals the chance.

Then actually run it

The most common failure is not a bad plan, it is a plan that gets filed and forgotten. If you set biweekly check-ins, hold them. If you promised a ride-along, do it. Running the plan honestly is what makes both outcomes fair: a real recovery you can trust, or an exit you know you gave every reasonable chance to avoid. Rules around documentation and termination vary by state and locality, so if the plan may lead to an exit, confirm your process with an employment attorney or your state labor office.

References

• Society for Human Resource Management (SHRM), performance improvement plan best practices
• U.S. Small Business Administration (SBA), managing and documenting employee performance
• See related: Documenting Performance So a Termination Holds Up; Put Someone on a Performance Plan or Not: Decision Tree