A Public-Space Maintenance SLA: Metrics, Penalties and Proof of Work

Start with service definitions, not promises

The most common failure in public-space contracts is language such as “maintenance as needed” or “keep the area tidy.” Such phrases cannot be measured, inspected or disputed. A service definition should state an objective outcome (for example, paths clear of litter and weeds), the frequency or response window attached to it, an acceptable tolerance, and what happens when weather or an event delays the work. The open-space specification system developed for parks in New Zealand and Australia illustrates the direction of travel: it assigns each maintenance activity one to five service-level indicators, with a star rating where five stars marks the highest level, so both client and contractor read the same requirement.

Industry frameworks for parks vendors converge on the same idea: replace subjective duty with a small set of measurable specifications covering frequency, quality standards, tolerances and contingencies. Before you write any number, split the territory into zones or park categories such as premium, high, medium and basic, because a central plaza and a neighbourhood green do not need the same regime. Documented levels of service also make regular performance audits possible later, because both sides know what “good” looks like in each zone.

• Vague: “mow the grass when needed.”
• Measurable: “mow to the agreed cut height at the zone’s scheduled frequency, and complete any rain-delayed cut within 24 hours of a cleared weather window.”
• Define who decides what “clean” means by referencing a numeric litter or condition scale in the specification.

Choose metrics that survive inspection

Quantifiable indicators tracked at least monthly form the backbone of the SLA. Common targets in facilities and public-works practice cluster around on-time completion of scheduled services (often 95% or higher), first-time fix rate (roughly 90%), response-time compliance, and an inspection-based quality score. The point of these numbers is that they are computable from records rather than from opinion, and each must carry an agreed method of counting so neither side can game the result.

Standard-setting bodies push in the same direction. The Italian standards body’s 2025 street-cleaning and urban waste management standard (UNI 11664-1) sets out how to define performance levels and conditions of acceptance so that quantitative and qualitative performance can be detected consistently, and how compensation for contractual non-compliance is formulated. Even formal national standards now expect municipalities to attach measurable acceptance criteria and compensation mechanics to cleaning and maintenance services rather than a vague duty of care.

• On-time completion: share of scheduled zone services finished on the agreed date.
• First-time fix: issues resolved on the first visit rather than after repeat call-outs.
• Response-time compliance: percentage of requests answered within the window set for their priority.
• Quality score: average of inspection ratings on a fixed scale, for example out of five.

Build a graded penalty and incentive ladder

A single fine for “failing to keep the area clean” is hard to apply fairly and invites disputes. A tiered structure lets consequences escalate with the severity of underperformance while rewarding strong delivery. A widely used shape has three tiers: minor non-compliance (for example 90–94% monthly compliance) triggers a written warning and a corrective plan; moderate non-compliance (85–89%) brings a modest percentage reduction in that month’s payment; and severe non-compliance below that threshold brings a larger reduction plus a formal contract review that may lead to termination. Good contracts add a positive incentive, such as a small bonus for sustained high compliance combined with high inspection scores, so the relationship is not purely punitive.

Whichever numbers you choose, the mechanism must be enforceable. The clause should say how compliance is computed each month, who certifies the data, and how credits or deductions are reconciled against invoices. Automated systems that clock vendor arrivals, tag photos and match invoices to completed work orders make penalties real rather than aspirational; without such a mechanism, a penalty clause is rarely invoked in practice.

• Tier 1 — minor: written warning plus corrective action plan.
• Tier 2 — moderate: a small percentage reduction of the monthly payment.
• Tier 3 — severe: a larger reduction plus contract review with possible termination.
• Positive incentive: a small bonus for sustained high compliance and high quality scores.

Schedule the seasons and set response windows

Public space is seasonal, and an SLA that ignores winter is unenforceable in a snow climate. Municipal standards illustrate the level of detail that works in practice: summer sweeping of intra-block driveways twice a week, snow clearing on a defined cadence when there is no fresh snowfall, snow removal to a storage site within a fixed number of days after a snowstorm, daily emptying of bins and a monthly washdown of those bins. The exact numbers vary by climate, zone and budget, but they show the granularity an SLA needs: a frequency for routine work and a deadline for response to events.

Response targets should be grouped by priority rather than covered by one blanket number. A typical ladder: critical safety hazards (broken glass, exposed wiring) get a response within hours and resolution the same day; high-priority equipment failures within a day or two; aesthetic issues such as overgrowth or graffiti within days; and routine scheduled work simply on its calendar date. Make the escalation path and the responsible party explicit for every level so there is no argument about who should have acted and when.

• Critical safety hazard: response in hours, resolution same day.
• High-priority equipment failure: response within a day, resolution within about 24 hours.
• Aesthetic issue (overgrowth, graffiti): resolution within a few days.
• Routine scheduled work: per the published zone calendar.

Prove the work, or it did not happen

The dispute that decides whether an SLA works is rarely about the KPI spreadsheet — it is about evidence. Require the service provider to keep dated service logs and, where practical, geo-tagged photo documentation of before and after states, and to deliver weekly status reports and monthly performance summaries. Require that spot inspections and complaints are recorded in two-sided acts or reports acknowledged by both parties, because a document the other side has accepted is far harder to challenge later.

Adopt a sampling-based inspection approach. Because no one can inspect every square metre daily, choose a defensible random or risk-weighted sample of zones each period, score them on the agreed scale, and accumulate those scores into the monthly quality metric. Publish the inspection calendar and criteria in the contract so the method cannot be accused of being arbitrary. A transparent method is what lets a client turn an honest audit into enforceable consequences rather than a personal grievance.

• Daily: provider service log with dates, zones, tasks and materials used.
• Per visit: timed before/after photo record where practical.
• Weekly: completed versus scheduled report with weather and exception notes.
• Monthly: two-sided inspection act and KPI summary used for invoicing.
• Quarterly: trend review and adjustments to zones or targets.

Keep the ladder fair and enforcement proportionate

Penalties are only credible if they are proportionate, documented and applied consistently. Courts and regulators push back against overreach: when a contract mixes a fixed penalty for any breach with a separate penalty for quality failures, a fine sized for a serious breach cannot simply be re-applied to a minor quality slip. Define precisely which clause applies to which kind of failure, and pair penalties with a written warning or a two-sided act so the provider knows what was found and when it must be corrected.

Finally, build forgiveness into the system. Many jurisdictions require or allow the client to write off penalties when the contract has been fully performed and the accumulated amounts are small relative to the contract value, and legitimate disruptions such as severe weather should be documented as exceptions rather than counted as failures. Apply the same discipline to the client: pay on verified evidence, keep the method transparent, and tighten targets only after the provider has had a fair chance to establish a realistic baseline. The goal is accountability that both sides can defend, not a fine machine.

• Qualify each clause: one penalty for a generic breach, a separate one for a quality failure.
• Pair every penalty with a written warning or a two-sided act.
• Define a write-off rule for small accumulated amounts on fully performed contracts.
• Document weather and force majeure as exceptions before they become disputes.

Practical asset

The checklist below works both before you sign or renew a maintenance agreement and at each review period. Mark every item yes or no: a “no” marks a gap you must close before the SLA can be enforced fairly.