Draft

Draft a performance improvement plan that holds up

Turns a manager's list of complaints into a PIP with dated facts, goals a third party could score, real support, and a routing flag when the timing looks bad.

About 15 minintermediateEmployment, In-house

Your prompt4,721 characters

Still to fill in: Employee and role, Performance concerns, Improvement goals, State and internal constraints

RoleYou are an employment lawyer who has read hundreds of PIPs and can tell within a page which ones the manager will defend line by line two years later. You replace adjectives with counts and dates, and you refuse to put a concern in the memo that the employee is hearing for the first time.What I needDraft the performance improvement plan for Employee and role, running 30 days, under the constraints in State and internal constraints. Flag anything in Recent protected activity that must be resolved before delivery.InputsEmployee: Employee and role Concerns: Performance concerns Goals: Improvement goals Duration: 30 days State and internal constraints: State and internal constraints Recent protected activity: Recent protected activityHow to work this1. Convert each item in Performance concerns into a fact (what happened, on which dates, against what standard) and name where it was previously raised, with the date. A concern that fails either test gets flagged for cutting, not written in anyway. 2. Test each goal in Improvement goals: could someone holding the data score it pass or fail without asking the manager? Rewrite the ones that fail, and flag any that cannot be met inside 30 days. 3. Make the support real: named training with a date, a named peer, a fixed cadence, the obstacle the company will remove. Support that will not be delivered is worse than none. 4. State consequences explicitly: what successful completion means, and that failure may result in termination. No implication, no euphemism. 5. Check State and internal constraints for constraints: a handbook procedure the company promised, a bargaining-agreement step, a statute protecting the conduct in Recent protected activity. If Recent protected activity names anything, open the memo with a routing flag: confirm a dated document showing the concerns predate the activity, comparable treatment of comparable performers, and goals no harsher than others in this org.Close with these four sections, every time, without being askedAssumptions I made. Which concerns I treated as previously communicated, that the goals sit within the employee's control, that the listed support exists. Mark each [verify] or [safe]. Where this is weakest. The two or three lines a plaintiff's lawyer would read aloud: an unmeasurable goal, a concern with no prior feedback, a target that depends on someone else. Name the section. What only you can decide. Options with tradeoffs, not flags. Run the full 30 days with every goal (more defensible, but it keeps a struggling employee in the seat longer) or narrow to the two goals that matter and decide sooner, which looks decisive to the business and pretextual to a jury. And whether to keep any goal that turns partly on factors the employee does not control. What would make this materially better. Ranked: the 1:1 notes for the dates cited, the last two reviews, the plans given to comparable performers here, or the handbook's own performance-management procedure.Output formatA confidential personnel-record memo: header, then numbered sections for purpose, role expectations, concerns, goals with target dates, support, cadence, mid-point review, successful completion, consequences, and an acknowledgment confirming receipt rather than agreement. Then a weekly check-in table, an end-of-plan outcome skeleton, and the routing flag on top if one is required.Never do this- If the memo would fit any underperforming employee anywhere, it is too generic. Every concern carries a date, a number, and a standard from my inputs. - No hedging filler and no HR euphemism. Cut "arguably," "needs to elevate," and "it depends." Do not tell me to consult employment counsel. I am writing this. - Never invent a date, a metric, an incident, or a prior conversation. A concern with no documented feedback gets marked [UNVERIFIED - confirm before delivery]. - Where you do not know whether a goal is achievable or a resource exists, say you do not know and ask. Do not smooth over the gap with confident prose. - Do not pad. Three concerns the manager can prove beat seven including a soft one. Length is not value; every weak item taints the strong ones.Before you answer- Is every concern tied to specific dates and a stated standard? - Could a third party with the data score each goal without asking the manager? - Does the memo contain any concern the employee has not previously heard? - If protected activity was named, is the routing flag at the very top?

Adds driver's-seat tunes: options instead of answers, questions before work, every citation flagged. Your values come with it.

Pressure-test it

Makes the AI switch hats and attack its own answer.

Picture the manager on the stand in a wrongful-termination trial, with the 1:1 notes and the comparable PIPs already in evidence and plaintiff's counsel reading this plan back line by line. Do that reading as the lawyer holding it. Name the three lines you would read aloud: an unmeasurable goal, a concern with no prior feedback, a target the employee could not control. Then rewrite each so the manager can defend it in one sentence without reaching for the file.

Go deeper

Pushes the work further once the basics are right.

How does a manager say all of this out loud, in a room, to someone who did not expect it? Write the delivery script for the PIP conversation: the opening, the walk-through of each concern, prompts that invite the employee's response without arguing, the support offer, and the close. Two pages at most, neutral in tone, with a note on what the manager should not say if the employee raises leave, a complaint, or an accommodation.

Before you run it

What to gather first

Each performance issue with dates and concrete examples
When each issue was previously raised, and where that is documented
Any leave, complaint, or accommodation request in the last several months
How comparable performers in the same org have been handled
What training, coverage, or tooling the company can actually provide

Watch for

A PIP delivered soon after leave, a complaint, or an accommodation request needs contemporaneous documentation that the performance issues came first. Confirm the document exists and is dated before the memo goes out.
Concerns the employee has never heard are the easiest thing for a plaintiff to attack. If prior feedback is not in the file, cut the concern or document it honestly as new.
Goals that depend on someone else's performance, on territory changes, or on a broken system are read as set up to fail. Check every goal for factors outside the employee's control.
Weekly check-ins have to actually happen and be written down. A plan with three of eight documented check-ins undermines the termination it was built to support.
If the plan ends in termination, the outcome memo must tie the failure to the original goals as written. Vague closeouts create more exposure than no plan at all.

What comes back

A confidential personnel-record memo with numbered sections from purpose through acknowledgment, covering concerns, measurable goals with target dates, support, cadence, mid-point review, and consequences, followed by a weekly check-in table, an end-of-plan outcome skeleton, and a legal-routing flag at the top where protected activity is present.

See an example of what you’ll get

MEMORANDUM - Confidential Personnel Record To: M. Aguilar | From: D. Park, Director of Sales | Date: April 27, 2026 | Re: Performance Improvement Plan LEGAL ROUTING REQUIRED: DO NOT DELIVER UNTIL HR AND LEGAL SIGN OFF. Employee returned from FMLA on January 15 and raised a pay-equity question with HR on February 28. Run every concern against those two dates before this goes anywhere: - *Forecast timeliness*: raised in 1:1s on 11/12 and 12/4, both before the January 15 return and before the February 28 question. Clean. - *Pipeline accuracy*: first raised on 1/22, which is a week after the FMLA return. It needs a pre-January 15 document (a CFO review note, a forecast-variance report, an email) or it comes out of this plan. - *Customer reliability*: the incidents are dated 1/8 and 2/14. The 2/14 incident postdates the FMLA return and every 1:1 on this list, so as of today she has never been told about it. Document a conversation first or cut the incident. - Confirm at least one comparable Sales Manager with similar metrics received a similar plan in the last 18 months, and that the goals and cadence here are no harsher than that comparator's. 1. Purpose. This plan identifies specific performance concerns, sets measurable goals, and provides the support and cadence needed to meet expectations over the next 60 days. 3. Performance concerns. 1. *Forecast timeliness.* Weekly forecast submitted more than 24 hours after the Monday 5:00 p.m. CT deadline in 6 of the last 10 reporting weeks (weeks of 10/13, 10/27, 11/3, 11/17, 12/1, 12/15). Raised in 1:1s on 11/12 and 12/4. 2. *Pipeline accuracy.* Five deals moved from Commit to Best Case during CFO review between November and January: Acme $310K, Helix $185K, Olive Tree $92K, Northshore $245K, Riverpoint $140K. Raised on 1/22. 3. *Customer reliability.* Two scheduled customer calls missed without advance notice or a reschedule (1/8 Acme, 2/14 Helix). Both customers escalated to the VP of Sales. [UNVERIFIED - confirm before delivery: no 1:1 in the record raises the 2/14 incident. If she has not been told about it, cut it or raise it in a documented conversation first.] 4. Improvement goals. (1) Forecast submitted by Monday 5:00 p.m. CT for eight consecutive weeks. (2) Commit-to-Best-Case reclassification under 10% at month-end review for May and June. (3) Zero unrescheduled customer no-shows through June 26. [Goal 4 in the input (Q2 attainment at or above 85% of quota) is flagged, not written in: Q2 closes June 30 and this plan ends June 26, so nobody can score it on the day the plan ends. Move the end date or drop the goal.] 5. Support. Weekly 1:1 with D. Park, Wednesdays 10:00 a.m. CT, focused on this plan. Two pipeline-review sessions with senior peer L. Brennan in May. Enrollment in the Forecasting Excellence workshop, May 14. IT to close the Salesforce report-build ticket she opened 4/2 by May 5. 7. Mid-point review. May 27, 2026: progress against each goal in writing, with support adjusted if needed. 9. If expectations are not met. If one or more goals are not met by June 26, 2026, employment may be terminated. The outcome will be documented in an end-of-plan memo tied to these goals as written. 10. Acknowledgment. Signing confirms receipt and review with your manager. It does not indicate agreement with the assessments. _______________ M. Aguilar Date: __________ Weekly check-in table. | Date | Goal 1 | Goal 2 | Goal 3 | Observations | Next steps | End-of-plan outcome skeleton. Restate each goal as written. Record the outcome with the underlying data. Conclude successful, not successful, or extended. Recommend the action and name who approved it. --- Assumptions I made. The 11/12 and 12/4 1:1 notes exist as written documents and say what D. Park remembers them saying [verify - the entire timing defense rests on those two files]. The six late-forecast weeks are recorded in the CRM with submission timestamps rather than reconstructed from memory [verify]. The Monday 5:00 p.m. CT deadline is a written, communicated standard and not an informal expectation [verify - if it appears nowhere in writing, concern 1 is an adjective]. L. Brennan has agreed to the two pipeline sessions and the Forecasting Excellence workshop actually runs May 14 [verify - support that does not arrive is worse than support never promised]. A 60-day plan with a written May 27 mid-point review satisfies the handbook's 30-day minimum and mid-point requirement [safe - Illinois, non-union, per your inputs]. Where this is weakest. Three lines. Concern 2: it was first raised on 1/22, a week after the FMLA return, and the five reclassifications are named but not dated individually. The first question in a deposition is which of the five happened after January 15, and this memo cannot answer it. Goal 2: "under 10% Commit-to-Best-Case reclassification" has no stated denominator: 10% of deals, of dollars, or of forecasted value? Someone holding the CRM data cannot score it without asking D. Park, which is exactly the test this plan is supposed to pass. Section 5: "IT to close the Salesforce report-build ticket she opened 4/2 by May 5" makes the company's own performance a precondition of hers; if IT misses that date, it is the first exhibit. What only you can decide. Deliver on the 60-day schedule as drafted, or hold until the pre-January documentation for concern 2 turns up. Delivering now keeps the business moving and stands on concern 1 alone, but the plan then carries a concern first raised a week after she returned from leave. That is the sentence plaintiff's counsel opens with. Holding costs three or four weeks and leaves a manager managing without a plan, and it buys a document where every concern predates every protected activity. Separately: three concerns or one. Concern 1 is provable to the day and nearly unattackable on its own; concerns 2 and 3 each carry a timing or notice problem, and a three-concern plan looks thorough to the business while giving the other side two things to pull on. And whether to keep any goal that turns partly on IT delivery, on a peer's availability, or on quarter-end timing. Cleaner to defend without them, more honest about the job with them. What would make this materially better. Ranked by impact. (1) The 1:1 notes for 11/12, 12/4, and 1/22 as dated written documents: they decide whether this plan is deliverable at all. (2) Anything dated before January 15 that raises pipeline accuracy; that single document moves concern 2 from a timing risk to a clean concern. (3) The last two PIPs issued to Sales Managers in this org, so the goals and cadence here can be shown to be no harsher. (4) The CRM submission timestamps behind the six late weeks. (5) The handbook's performance-management section verbatim, to confirm the May 27 review satisfies it.

Why this prompt is built the way it is

## Framework 1. **Facts, not adjectives.** "Submitted the forecast more than 24 hours late in six of ten weeks" beats "lacks attention to detail." 2. **No surprises.** Every concern in the plan must trace to feedback the employee already received, with the date of that conversation. 3. **Goals a stranger could score.** If someone with the underlying data cannot confirm pass or fail without asking the manager, the goal is not measurable. 4. **Achievable inside the period.** Goals set to fail are pretext evidence. So is a period too short to show improvement. 5. **Support that will actually be delivered.** Named training, a named peer, a fixed meeting cadence, and the obstacle the company will remove. 6. **Explicit consequences.** What successful completion means, and that failure may result in termination. No euphemism. 7. **Route it when the timing is bad.** Recent leave, complaint, accommodation request, or charge means the file must show the concerns predate the activity before the plan is delivered. 8. **Neutral tone.** Not a confession from the manager and not a termination notice in disguise.