Two operators, same phone. The first approves every request the second it buzzes, and by the tenth of the month has green-lit $600 of can liners for a building that consumes about $200 worth. The second lets requests pile up until Friday paperwork time, and on Wednesday night a crew lead pays $9 a roll for bath tissue at the grocery store on the way to the site.
Both of them think the problem is approval speed. It is not. The problem is that they are making a discretionary decision about a $14 case of liners at 9:40 p.m. on a phone screen the size of a playing card.
The workable rule for most janitorial operations: auto-approve any request under a set dollar threshold from a pre-approved product list, require a photo and a one-line reason above it, and commit to answering every request within one shift. Approve from the phone browser. Decide the policy at the desk.
What follows is the decision framework: the axes that actually determine which approval model you should run, a rule for each axis, and a straight recommendation for four common operation profiles.
What a supply request must contain before you can approve it from a phone
Most mobile approval pain is really data pain. If a request arrives as "need more soap," you cannot approve it responsibly from anywhere, including your desk.
Standardize the request format first. Five fields is enough, and all five fit on one screen.
The five-field supply request
- Location and closet: Which building, and which storage room inside it. Multi-tenant sites often have two or three.
- Item, exactly as it appears on your order sheet: "Jumbo jr. 2-ply tissue, 12 rolls per case," not "toilet paper."
- Quantity requested and quantity currently on hand: On hand is the field that catches over-ordering. Without it you are guessing.
- Photo of the shelf: One picture ends 90 percent of the follow-up questions. It also documents the closet condition.
- Need-by date: "Tonight" and "next delivery" are different decisions. Make the requester say which.
Once those five fields are mandatory, mobile approval stops being a judgment call and becomes a two-second check: does on-hand plus requested land near the par level for that closet? If yes, approve. If no, ask one question.
Should you auto-approve small supply requests? Run the arithmetic
Reviewing a request is not free. Assume it takes two minutes of real attention including the context switch away from whatever you were doing. If your own time is worth $50 an hour fully loaded, that review costs about $1.67 in owner time.
Now put that next to the item. A case of 33-gallon liners or two gallons of neutral cleaner is a low-dollar, high-frequency purchase. Spending $1.67 of attention plus a possible overnight delay to scrutinize a $14 order is a losing trade, every time.
A defensible way to set the threshold: pick the dollar figure where the item cost is roughly ten times your review cost, then sanity-check it against your monthly supply budget per site. With the $50 per hour assumption above, that lands somewhere near $20 in owner time value, and most operators end up setting the practical auto-approve line between $50 and $150 per request.
Below the line, the control is not your approval. The control is the catalog and the par level. Above the line, you actually read it.
How to set a par level so most requests never happen
The best supply request is the one nobody had to send. Par levels do that work, and they take one month of counting to establish.
Count the closet on the first of the month, count again on the first of the next month, and add in everything you delivered between. Consumption equals starting count plus deliveries minus ending count.
Then use this:
- Par level = (average weekly usage x weeks between deliveries) + one week of usage as safety stock
- Reorder point = one week of usage + whatever your distributor's lead time consumes
Worked example. Northgate Medical Plaza is 42,000 sq ft, cleaned five nights a week, and you deliver supplies every two weeks. Your month of counting shows 84 rolls of jumbo junior bath tissue consumed, so 21 rolls per week.
Par = (21 x 2) + 21 = 63 rolls. Reorder point sits around 25 to 30 rolls. Now any request that arrives while the closet holds 45 rolls is not a supply need, it is a habit, and you can decline it in one thumb-tap with a canned reply.
The numbers above are illustrative. Yours will differ by fixture type, traffic and dispenser style, which is exactly why you count your own buildings instead of borrowing someone else's benchmark.
Which supply approval model fits your operation
There are four models actually in use. The axes below decide between them. Read the row that matters most in your business and let it break the tie.
| Deciding axis | Approve every request | Threshold auto-approve | Par-level standing order | Delegate to site lead |
|---|---|---|---|---|
| Requests per week you can realistically read | Under 10 | 10 to 40 | Any volume | 40+ |
| Number of accounts | 1 to 5 | 5 to 20 | Stable, recurring sites | 20+ or multi-city |
| Product mix | Anything goes | Fixed catalog required | Fixed catalog required | Fixed catalog plus spend cap |
| Cost of a stockout | High risk of delay | Low | Lowest | Low |
| Shrinkage exposure | Lowest | Moderate, needs month-end check | Low, closet is counted | Highest without spend caps |
| Chemicals and anything with an SDS | You control it | Exclude from auto-approval | Pre-cleared products only | Exclude unless lead is trained |
| Owner time per week | Highest | Moderate | Lowest ongoing, high setup | Low |
| Audit trail for client-billable consumables | Strong if logged | Strong | Weakest at the request level | Weak without receipts |
| Works when you are on a job site | Poorly | Well | Not applicable | Well |
Under 10 cleaners, 1 to 5 accounts: approve every request, but standardize the format
At this size you know every closet personally and the request volume is small enough to read on your phone between stops. Do not build machinery you do not need yet.
What you do need is the five-field format and a written response window. The failure mode at this size is not overspending, it is a request sitting unread for four days while a cleaner improvises.
10 to 20 cleaners across 8 to 20 accounts: threshold auto-approve, hard catalog
This is where the mobile approval load becomes real and where most operators lose control. Commit to a threshold, publish a one-page approved product list, and auto-approve anything on the list under the line.
Everything above the line, plus every chemical, plus every piece of equipment, comes to you. That is usually two or three decisions a week instead of thirty.
Stable recurring accounts with predictable consumption: par-level standing orders
If a building has cleaned the same square footage on the same frequency for six months, requests are the wrong mechanism entirely. Set the par, put the delivery on the schedule, and handle only exceptions.
Exceptions are real: a burst pipe, a tenant move-in, a flu week that empties the sanitizer. Those still come to your phone, and now you have time to read them.
20+ cleaners or multiple cities: delegate to site leads with a spend cap
You cannot be the bottleneck for 60 closets in four metros. Push approval authority down to the person who can walk to the shelf, and cap it: a per-request limit and a monthly limit per site.
Pair delegation with reconciliation, not with trust alone. A monthly comparison of approved requests against distributor invoices and closet counts is the control that makes delegation safe.
How fast do you have to answer a supply request?
Set a service level for yourself, because your crews already have an informal one and it is shaping their behavior.
A practical standard: requests marked "next delivery" get an answer within 24 hours. Requests marked "tonight" get an answer before the next scheduled shift at that site, or a phone call.
Then batch. Two review windows a day, morning and early evening, beats reacting to every notification. The evening window matters most, because that is when the crew is standing in the closet and can still act on your answer.
Consider what a missed answer actually costs. If a cleaner makes an unplanned supply run of 40 minutes at a $22 per hour fully loaded rate, that is roughly $15 of labor plus mileage plus retail pricing on the item. Two of those a month across ten sites is not a rounding error. Use your own loaded rate and the current wage figures for your metro rather than the illustrative numbers here.
Which requests should never be auto-approved
Three categories belong in your hands regardless of dollar amount.
- Any chemical not already on your approved list: Under OSHA's Hazard Communication Standard, 29 CFR 1910.1200, you are responsible for maintaining safety data sheets and training employees on the hazardous chemicals in their workplace. A cleaner buying an unfamiliar degreaser creates a compliance obligation you did not plan for.
- Equipment and repairs: A vacuum belt is a supply. A vacuum is a capital decision with a warranty question attached.
- Anything you bill through to the client: If consumables are pass-through under the contract, the request is also an invoice line. It needs the same scrutiny as a change order.
How to catch over-ordering without reading every request
Auto-approval only works if something checks it after the fact. Do this once a month, and it takes about 20 minutes per account.
- Pull the approved requests for the month by site, with quantities.
- Pull the distributor invoices for the same period.
- Have the supervisor photograph and count each closet on the same date every month.
- Compare consumption per site against square footage and occupancy. Flag any site whose supply cost per square foot moved more than a little from the trailing three months.
- Investigate the flagged sites only. Usually it is a dispenser problem, a new tenant, or product being carried to another building.
Track supply cost per cleanable square foot per month as your one number. It normalizes across buildings of different sizes and makes an outlier obvious without any judgment about individual requests.
Frequently asked questions
Can I approve supply requests from a phone browser, or do I need a native app?
A mobile browser is sufficient for approvals if the request form is short and the approve action is a single tap. What matters is that the record lives in one system with a timestamp and the requester's name, not whether it arrived through an app icon. Many cleaning operators run their whole approval flow through a browser link saved to the phone home screen.
What dollar threshold should I set for auto-approval?
Start by valuing your review time, then set the line where the item cost is roughly ten times that. For most small and mid-size janitorial companies the practical range is $50 to $150 per request. Set it lower for a new crew member during their first 90 days, then raise it once you have seen their ordering pattern for a couple of months.
Should client-supplied consumables be approved the same way as mine?
No. If the client owns the supplies, your job is notification and documentation, not purchasing. Route those requests to the facility contact with the photo attached and log the date you sent it. That log protects you when a tenant complains about empty dispensers and the order was never placed on the client's side.
What do I do when a cleaner buys supplies out of pocket without approval?
Reimburse it this time if the need was real, then fix the cause. Write a reimbursement rule: itemized receipt, photo of the empty shelf, submitted within 48 hours, and a stated ceiling per incident. Repeated out-of-pocket buying at one site almost always means the par level is too low or your response window is too slow.
How do I stop supply requests from getting lost in group texts?
Group texts have no status field, so nothing is ever officially open or closed. Move requests to one channel that produces a record with a site, an item, a quantity and an approval timestamp. Keep the group chat for coordination and tell crews plainly that a supply request sent by text will not be treated as submitted.
How CleanTrack360 fits
CleanTrack360 includes supply requests alongside scheduling, geofenced GPS clock-in and clock-out, quality inspections with photo evidence, and location-based team messaging, all running in the phone browser. There is no published mobile app yet, so crews submit from their phone browser and you review from yours between site visits. Reports export to CSV, which is what you want for the monthly reconciliation described above, and the client dashboard gives facility contacts visibility into schedules, inspection reports and service requests.
Plans are priced per plan rather than per user: Starter at $99 per month for up to 5 team members, Pro at $199 per month for up to 20, and Business at $249 per month for up to 50. There is a 14-day free trial with no credit card required, which is enough time to count one month of closet usage and set your first par levels.