Vendor Performance Improvement Plans That Work

A vendor improvement plan is a short, signed document that converts a performance complaint into a schedule with numbers in it. It names what failed, what success looks like, how it will be measured, who checks, and what happens if the measure is not met by a stated date.

Most plans fail for one of two reasons. Either they are written as a warning letter with no measurable target, or they are so long and legalistic that nobody works from them. A useful plan is one to two pages and fits on the clipboard of the supervisor who has to deliver it. It sits between the review and the formal notice in the wider property management vendor management lifecycle.

Key takeaways

• Write a plan when there is a pattern, not a single miss. One bad night is a conversation; a pattern is a plan.
• Every target must be measurable with the inspections you already run. "Improve washroom standards" is not a target.
• Set a 30-day plan with weekly checkpoints. Long plans lose momentum and give the vendor a place to hide.
• Say what happens on the review date, in the contract's own language — further notice, set-off, or termination for cause.
• Both parties sign it, including the vendor's operations lead, not just the site supervisor.
• Inspect the same areas the same way before and after, or the plan's evidence is worthless.

When is an improvement plan the right step?

An improvement plan belongs in the middle of the escalation ladder, after a documented conversation and before a formal notice. It is the right step when the vendor has acknowledged the problem but the pattern has continued, or when the score gap is clear enough that a structured recovery is worth one attempt.

It is not the right step in three situations:

• A single isolated miss. Fix it with a written note and move on.
• A critical safety or hygiene breach. That is a report-and-stop event, handled immediately and separately.
• A capacity problem. If a vendor genuinely cannot staff the building, no plan will change the arithmetic. Reduce scope, add a vendor, or retender.

If you are unsure whether you are at the plan stage, the sequence is set out in what to do when a vendor's performance drops below standard.

What an improvement plan must contain

Eight things, in this order. Anything that does not fit on two pages is not a plan, it is a report.

1. Parties and dates. Building, contract reference, plan start date, review date, and the named contacts.
2. The findings. Each failing item, quoted against the specification clause it breaches.
3. The evidence. Inspection dates, area scores, photo references and repeat counts.
4. The targets. A measurable result per finding, expressed in the same scoring you already use.
5. The measurement method. Who inspects, how often, in which areas, and with what checklist.
6. The checkpoints. Weekly dates for the first month, with a named attendee per side.
7. The support you will provide. Access changes, a meeting, schedule adjustments you control.
8. Non-negotiables outside the plan. Chemical labelling, wet floor signage and PPE compliance are pass or fail on the day — critical.
9. The consequence. What happens on the review date if the target is not met.

Keep the tone neutral. A plan that reads as a punishment gets defended rather than implemented.

Write targets that cannot be argued with

The quality of the plan lives in the targets. Use the scoring and reporting you already produce, because a target the vendor has never seen before is a target they will dispute.

Three rules make targets stick. Express them in your existing scoring scale, so both sides already understand what a 3.5 means. Attach a count and a period to every one. And keep the number of targets small — five or six at most. A plan with fifteen targets has none.

CCOHS inspection guidance recommends recording a responsible person, a due date and an action status for every finding, and copying unfinished items onto the next report until they close. Borrow that discipline for the plan: an open item with a name and a date is harder to ignore than a paragraph of concern.

The 30-day structure

Thirty days is the right length for most cleaning contract plans. It is long enough for a trend to appear across four or five inspections and short enough that attention does not drift.

Week 1: stabilise. The immediate failures — dispensers, bins, entrance floors, washroom fixtures — should be visibly fixed within seven days. Inspect at the end of week one and share the result the same day.

Week 2: prove consistency. The same items, the same areas, same method. A single good week is a resourcing push; two good weeks is a routine.

Week 3: extend to periodic work. Bring the neglected periodic tasks into the plan — extraction, high dusting, floor care — and confirm scheduling, not just intent.

Week 4: full review. Run a complete scored inspection across every area, compare to the baseline, and hold the review meeting within three business days.

For a multi-site contract, run the same plan across all affected sites with a separate checkpoint per building. Consistent scoring across a portfolio is what makes the results comparable, which the portfolio vendor management guide covers in detail.

Run the review meeting properly

The review meeting is short and evidence-led. Sit down with the inspection reports on the table — printed, not described — and work through the targets one by one.

• State the result per target, pass or fail, with the underlying scores.
• Ask one question per failed target: what specifically will change in the routine?
• Agree actions with a name and a date, not intentions.
• Decide the next step, whether that is closing the plan, extending it once, or moving to formal notice.

Do not renegotiate the targets in the review meeting. If a target turned out to be genuinely unmeasurable, that is a lesson for the next plan, and you can record it as an amendment signed by both parties.

If the plan succeeded, say so in writing and close it. A vendor who recovers and gets no acknowledgement learns that recovering is not worth the effort.

What happens if the plan fails

Meet the consequence you wrote, or the plan has taught the vendor that your deadlines are decorative. Depending on your contract, the options are:

• Issue a formal notice with a cure period, and continue inspecting at the same frequency.
• Apply contractual remedies — service credits, a set-off against the invoice, or a requirement for additional supervision at the vendor's cost.
• Escalate the resource question with the vendor's operations director: is this site resourced to the specification at all?
• Begin the exit process, including the records and access steps in the vendor offboarding process.

Not regulatory advice. Contractual remedies, set-off and termination rights depend on your agreement and on the law of your province or territory. Confirm your position before exercising any of them.

Common ways improvement plans fail

• No baseline. Without a scored starting point, nobody can prove the plan worked or failed.
• Targets written by the manager and never discussed. The vendor's supervisor receives a document they had no part in and quietly ignores it.
• Weekly checkpoints that get cancelled. Two cancellations and the plan is dead.
• Changing the inspection method mid-plan. Different areas or a different checklist make the comparison meaningless.
• No consequence, or a consequence you never intended to impose. Both destroy credibility.
• Running a plan instead of fixing the scope. If the specification no longer matches the building, write a variation rather than a plan.

Frequently asked questions

How long should a vendor improvement plan last?

Thirty days for most single-site cleaning contracts, with weekly checkpoints. Larger or multi-site contracts sometimes need 60 or 90 days, but only if you can sustain weekly attention for that long. A plan nobody checks is worse than no plan.

Should the improvement plan be a formal warning?

No. Keep it as a management document you both sign. Formal warnings and notices to perform are separate contractual instruments with legal consequences, and mixing the two makes every subsequent conversation a legal negotiation.

What if the vendor refuses to sign the plan?

Record the refusal in writing on the day, send the plan by email, and state that you will inspect against it regardless. A refusal to sign is itself useful evidence of the vendor's engagement with the contract.

Can I use an improvement plan for a subcontractor?

You can require it through your main vendor. Your contract is with the vendor, not their subcontractor, so the plan should name the vendor as the responsible party and require them to manage their own subcontractor's performance.

What score improvement should I expect in 30 days?

Expect the failing areas to recover to your contract's pass threshold, not to perfection. If a site is scoring 2 on a 5-point scale, a realistic 30-day target is 3.5 with no critical items outstanding. Depth takes longer than discipline.

Set the baseline before you write the plan

An improvement plan is only as good as the evidence behind it, and that evidence comes from inspecting the same areas the same way, week after week. Without a scored baseline, the review meeting becomes an argument about impressions.

NeatScore gives property managers that baseline. Build area-specific checklists with critical items flagged, score each area on the APPA 1–5 scale, attach photos per item, and run inspections offline in basements and parkades where the signal drops. Weighted scoring and deficiency tracking show which targets moved, and dated branded reports give both sides the same document at every checkpoint. Try it free for 14 days, or see pricing.