This guide is for facility managers, building service contractors, and procurement officers who are responsible for keeping consumables in stock without tying up cash in slow-moving inventory. If you’re running three or more buildings, or if you’ve ever experienced a mid-week stockout on paper towels at the same time you discovered two pallets of hand soap sitting past their print date, this is for you.
The goal is to build a functional par-level system — not a polished one, not an expensive one — that prevents both stockouts and the accumulation of dead inventory. Most facilities think they already run a par-level system. Most of them are actually running an emergency-reorder system in a trench coat.
What a Par Level Actually Is (and What It Isn’t)
A par level is not “we usually keep a few cases on the shelf.” That is a vague intention, not a system.
A proper par-level system defines, for every active SKU:
- Minimum stock (par): the quantity below which you must place a reorder
- Maximum stock: the upper limit you will not exceed during a normal order cycle
- Reorder point (ROP): the specific quantity that triggers the reorder, calculated from lead time and usage
These three numbers are derived from usage data — not from memory, not from the distributor’s suggestion, and not from “we bought that last year and it worked out fine.”
The system is only as good as the usage data behind it. If you do not track usage, par levels are theater.
The Math Behind Par Levels
This is not complex, but it has to be done correctly for each SKU.
Average daily usage (ADU)
Count how much of a given SKU was consumed over a defined period — 30 days is workable, 60 days is better, 90 days captures seasonal variability. Divide by the number of days.
ADU = Total units consumed ÷ Days in period
Example: 180 rolls of paper towels consumed over 60 days → ADU = 3 rolls/day
Lead time (LT)
Days from the moment you place an order to the moment the product is received, shelved, and accessible. Do not use the distributor’s quoted transit time. Measure your actual order-to-shelf time, including your own receiving and stocking process. For most local distributors, this runs 1–3 business days. For specialty or imported products, it can be 7–14 days or longer.
Safety stock (SS)
The buffer that absorbs variability in both usage and lead time. There are two approaches:
Rough cut (adequate for most janitorial programs):
SS = 25–50% × (LT × ADU)
For a stable consumable with predictable demand, use 25%. For anything with usage spikes or an unreliable supplier, use 50%.
Statistical method (better for high-value SKUs or tight storage):
SS = 1.5–2.0 × σ (standard deviation of daily usage) × √LT
This is more precise but requires at least 30 days of daily usage data. For most facilities, the rough-cut method is close enough.
Reorder point (ROP)
ROP = (LT × ADU) + SS
This is the minimum quantity in stock that should trigger a reorder. Not “we’re getting low.” A specific number.
Maximum stock
Max = ROP + Typical order quantity (or 2–4 weeks of supply, whichever is lower given storage constraints)
Worked Example: Paper Towels in a Class A Office Building
- ADU: 3 rolls/day (tracked over 60 days)
- Lead time: 2 business days (measured actual)
- Safety stock (rough cut, 25%): 0.25 × (2 × 3) = 1.5 → round to 2 rolls
- ROP: (2 × 3) + 2 = 8 rolls
- Order quantity: 24 rolls (half-case)
- Max stock: 8 + 24 = 32 rolls
That facility should have bin labels reading MIN: 8 / MAX: 32. When a count shows 8 rolls or fewer, reorder is triggered.
The SKU Classification Problem
Not every product behaves the same way. Applying identical par logic to all SKUs is one of the most common — and costly — mistakes in janitorial procurement.
| Category | Usage Pattern | Shelf Life | Par Strategy |
|---|---|---|---|
| Toilet paper, paper towels | Steady, predictable | 2+ years | Standard ROP formula; stable ADU |
| Hand soap (foaming concentrate) | Steady with slight seasonal variation | 18–24 months | Standard; slightly wider SS in flu season |
| Neutral floor cleaner (concentrate) | Steady | 12–24 months | Standard; watch dispensing system consistency |
| Quat disinfectant concentrate | Steady to seasonal | 12–24 months | Standard; enforce FIFO strictly |
| Enzymatic cleaner/drain treatment | Moderate, episodic | 6–12 months | Tighter max; order more frequently in smaller quantity |
| Peracetic acid sanitizer | Low volume, project-driven | 6–12 months | Near-zero safety stock; order on demand |
| Ice melt (calcium chloride, magnesium chloride) | Extreme seasonal spike | 2+ years if sealed | Seasonal par only; zero standing inventory April–October |
| Floor stripper (high-alkaline) | Seasonal project peaks | 12–24 months | Project-based ordering; small standing stock |
| Graffiti remover, specialty solvents | Very low, episodic | Varies | Minimal standing stock; do not bulk buy |
The reason this table matters: buying 18 months of enzymatic cleaner at a favorable bulk price sounds smart until month 8, when you discover the last four cases are past the manufacturer’s use-by date. Enzymatic products depend on live bacterial or enzyme activity. They degrade. A quat disinfectant that’s 14 months old and stored in a hot storage room may have lost 20–30% of its active concentration.
Why Facilities Get This Wrong
Anecdotal usage estimates instead of counted data
“We go through about two cases a week” is not usage data. It is a guess. Someone walked off with a partial case to another floor, supplies were used for a special event, or a leaky dispenser was replaced without logging the wasted product. None of that shows up in the “we think” number.
Counted data means physically counting on a consistent schedule, recording the number, and comparing it to the prior count.
Not accounting for shelf life
A stockout of paper towels is a nuisance. A stockout of a specialty stripper means your weekend project doesn’t happen. A stockout of an NSF-registered food-contact sanitizer in a food-service facility is a food safety incident with regulatory implications.
But the inverse is also true. Buying 24 months of an enzymatic drain treatment when the product has a 9-month shelf life is not inventory management — it is waste management scheduled 9 months in advance.
Treating the ROP as a suggestion
The reorder point is a number that triggers a specific action: place the order. It is not a reminder to think about placing the order. If the decision-maker sees 10 rolls of paper towels against an ROP of 8 and decides “we’ll be fine until Friday,” the system has failed.
No actual usage capture
Consumables disappear in facilities without records. A custodian pulls supplies for an urgent spill cleanup; the quantity isn’t logged. A paper towel dispenser jams and a case is wasted; no one records it. Over weeks, your usage data drifts from reality, and your par levels drift with it.
Setting Up the System Without Enterprise Software
You do not need a CMMS or an ERP to run this. The following works for a 1–5 building operation.
Bin labels
Every supply location gets a label with four fields: SKU name, unit of measure, MIN quantity, MAX quantity. Color coding is optional but useful — some facilities paint a red line on the inside of the bin corresponding to the minimum level.
Weekly count protocol
Assign the count to a specific person on a specific day. Thursday morning count, order placed Thursday afternoon, delivery Friday or Monday. The consistency matters more than the day. A weekly count cadence works for most SKUs. Daily for restroom paper products in high-traffic facilities.
Usage tracking (minimal version)
A shared Google Sheet or Excel file with columns: Date, SKU, Location, Opening Count, Closing Count, Units Consumed. One row per SKU per week per location. This is an hour of work per week for a three-building operation. The data is what lets you calculate a real ADU.
Sample usage tracking sheet structure:
| Date | SKU | Location | Open | Close | Used |
|------------|------------------------|---------------|------|-------|------|
| 2024-11-04 | Paper Towels, M-Fold | Bldg A, 3F | 48 | 22 | 26 |
| 2024-11-04 | Paper Towels, M-Fold | Bldg A, 6F | 24 | 11 | 13 |
| 2024-11-04 | Quat Disinfectant Conc | Janitor Closet| 6 | 4 | 2 |
Order cadence
Set order days and hold to them. Ad-hoc reordering creates two problems: you either catch stockouts late (after the ROP was crossed) or you over-order because you’re buying to feel safe rather than buying to a calculated quantity. A set cadence — every Tuesday, first Monday of the month — makes reordering mechanical rather than reactive.
Seasonal Flexing
Par levels are not static. The same facility has different usage patterns in January versus July.
Ice melt: Zero standing inventory from April through October unless you have free, dry storage. If you’re buying calcium chloride or magnesium chloride in October “just in case,” you are paying storage cost (or opportunity cost) for 6+ months on a product that deteriorates when exposed to humidity. Buy on a seasonal curve: small quantity in October, reorder based on weather forecasts.
Restroom consumables during peak occupancy: A 200,000 sq ft office building at 40% occupancy in August needs different par levels than the same building at 95% occupancy in October. Track occupancy data alongside usage data if your buildings have measurable occupancy variation.
Summer floor maintenance: Strip-and-refinish projects, concrete sealing, pressure washing programs — these create demand spikes for strippers, sealers, and high-alkaline degreasers that don’t show up in your baseline ADU. Treat these as project-based purchases separate from standing par-level inventory.
Revisit par levels at least quarterly. Revise them when you’ve collected 30+ days of new data that materially differs from the baseline.
The Storage Constraint Is Real
In most urban office facilities, storage space is the binding constraint. A 1.5M sq ft Class A portfolio across three buildings in a dense urban market may have 400–600 sq ft of total janitorial storage across all locations combined. That is not warehouse space. Chemical storage must comply with OSHA 29 CFR 1910.1200 requirements for hazardous materials and with your local fire code.
Chemical compatibility in storage:
- Acids and bases must be segregated (acid toilet bowl cleaners stored with an alkaline floor stripper creates a fume and spill hazard)
- Oxidizers (peracetic acid, hydrogen peroxide-based products, some pool chemicals if any) must be stored away from flammables and organics
- Concentrated products generally have tighter storage requirements than diluted RTU formulas
Vertical storage (racking) recovers square footage. Accessibility is the constraint: if the right-sized pallet rack means your custodians can’t safely reach the top shelf, it’s not practical.
The point for par-level setting: your max stock quantity may be limited not by carrying cost or shelf life, but by physical storage capacity per location. Build that ceiling into your max-stock calculation.
Named Scenario: 1.5M Sq Ft Class A Office Portfolio (3 Buildings)
A property management company operates three Class A office towers totaling 1.5 million square feet across one metropolitan market. They have no formal par-level system. Each building manager reorders when they run low, with no shared visibility. The result: Building A has $14,000 in paper products in storage; Building B ran out of toilet tissue twice last quarter; Building C has 40 gallons of a floor finish they no longer use after switching dispensers.
Phase 1 (Days 1–30): Conduct a full inventory count at all three buildings. Categorize SKUs as high-use/stable (paper products, trash liners, neutral cleaner), moderate-use/stable (quat concentrate, hand soap), and low-use/variable (specialty chemistries, seasonal products). Set initial par levels using rough-cut safety stock (25% for stable SKUs, 50% for variable).
Phase 2 (Days 30–90): Run weekly usage tracking at all three buildings using a shared spreadsheet. Calculate actual ADU from counted data. Compare to initial estimates. In this scenario, the initial ADU estimate for paper towels at Building A was 8 rolls/day; actual tracked usage showed 11 rolls/day (a tenant event space was not accounted for). Par levels adjusted upward.
Phase 3 (Day 90 onward): Revise all par levels based on 60 days of usage data. Set seasonal par levels for ice melt and floor project supplies. Establish a single weekly order day for all three buildings with one distributor. Total SKU count rationalized from 87 active SKUs to 54 after identifying duplicates and discontinuing slow-movers.
Outcome: One stockout in 6 months post-implementation (a specialty product with a 7-day lead time ordered too late before a holiday). No dead inventory written off in that period.
Vendor-Managed Inventory: When It Helps, When It Doesn’t
Some distributors offer vendor-managed inventory (VMI) — they monitor your stock levels and replenish automatically. This sounds attractive. It sometimes is.
VMI works when the distributor has strong product and location coverage, when your usage is stable enough to make their replenishment model accurate, and when you have a multi-year relationship with leverage to negotiate service terms.
VMI is a procurement disadvantage when you surrender your order data without getting something concrete back. If the distributor sets your par levels, controls the order quantity, and determines replenishment timing, you have lost visibility into your own usage data, and you’ve lost the volume leverage that comes from competitive bidding. The supplier now knows exactly how much you use and when you reorder. That is valuable intelligence, and you’re giving it away for free.
A middle path: run your own par-level tracking and share order timing with your distributor — but set your own quantities and retain the right to rebid annually.
Common Mistakes
- Setting par levels and never revisiting them. A 90-day-old par level based on summer data will be wrong by December.
- Ignoring shelf life. Check manufacturer shelf-life specs for every chemistry SKU. Record the manufactured/fill date on incoming inventory. Enforce FIFO at the storage level.
- Bulk-buying ice melt in October. The “just in case” rationale does not account for storage cost or product quality degradation in a non-climate-controlled space.
- Not differentiating high-use from slow-mover SKUs. Applying the same par logic to paper towels and peracetic acid sanitizer is not a system; it is a template imposed on reality without thought.
- Using the ROP as a guideline rather than a trigger. The system works when the trigger is respected. Discretion at the moment of reorder breaks the system.
Printable Par-Level Setup Worksheet
PAR-LEVEL SETUP WORKSHEET — [Building/Site Name] _______________
Tracking period: ________________ to ________________
For each active SKU:
SKU Name / Description: _______________________
Unit of measure (case, roll, gallon, etc.): _______________________
Shelf life (months): _______________________
Step 1 — Usage tracking
Total units consumed during tracking period: _______
Days in tracking period: _______
ADU (consumed ÷ days): _______
Step 2 — Lead time
Average days from order to shelf (measured): _______
Step 3 — Safety stock
[ ] Rough cut (25% of LT × ADU for stable SKU): _______
[ ] Rough cut (50% of LT × ADU for variable SKU): _______
Step 4 — Reorder point
ROP = (LT × ADU) + SS = _______
Step 5 — Maximum stock
Typical order quantity: _______
Max = ROP + Order qty = _______
Storage capacity maximum: _______ ← use lower of the two
Step 6 — Seasonal adjustment?
[ ] Yes — seasonal par applies from _______ to _______
Seasonal ADU: _______ → Recalculate ROP: _______
[ ] No
Label: MIN _______ / MAX _______
Review date (90 days): _______________________
Assigned counter: _______________________
Order day: _______________________