A machine can look perfect from the outside and still be losing money. Glass clean, coils full, every bag facing forward. Then you pull the sales report and find that eleven of the twenty eight slots have not moved a unit in six weeks, and that the inventory sitting in them is quietly approaching its date code.
That is not a stocking problem. It is a planning problem, and vending machine inventory management is the discipline that fixes it. Four jobs, done in order, decide whether a route earns or just stays occupied.
Four jobs inside vending machine inventory management
The first job is the layout. Which product sits in which slot, at what price, with how many facings. The second is the fill level, meaning how many units each slot should hold when you leave. The third is rotation, so the oldest unit sells first. The fourth is reconciliation, comparing what you loaded against what sold.
Each one feeds the next. The layout sets capacity, capacity sets the fill level, the fill level builds the pick list, and the pick list is what the driver carries through the door. Skip the layout and the other three become weekly guesswork.
Build the planogram from the machine, not from a catalogue
A planogram is the backbone of vending machine inventory management, and it is a slot by slot map. For each position it records the slot ID, the SKU and its barcode, the number of facings, the fill target and the vend price. That sounds bureaucratic until you watch two drivers stock the same machine differently and try to compare their sales.
Start with the machine itself. Count the trays and the columns, note which columns are wide or double spirals, and check the vertical clearance on each shelf. Then assign product families to zones so a customer can scan the glass in a few seconds. Salty snacks with salty snacks, drinks grouped by type.
Within a zone, separate packages that look alike. Two brands of plain chips in identical red bags read as one product from two feet away, and the customer buys neither. The trade calls this breaking the colour block, and it costs nothing to do.
Weight and shape decide the rest. Heavy bottles belong on the bottom trays of a combo machine, where the coil pitch and the delivery chute can carry them. Tall light bags need vertical clearance. Narrow bars and gum suit high count spirals. A jam in a premium slot is worse than an empty one, because it takes the machine offline for everything behind it.
The last rule is the one most operators get wrong. Your fastest sellers need more than one column. A single column of a top seller sells out first, and the day it empties is the day you decide to visit, whether or not the other twenty seven slots needed anything.
Par levels, and why par is not capacity
Capacity is the most a slot holds without jamming. Par is what you fill it to on a normal visit. They are not the same number, and treating them as the same number is how slow sellers end up with six weeks of stock aging in plain sight.
A workable starting rule is par at one and a half times weekly sales, rounded to what the coil holds. If a slot sells ten units a week and you service weekly, par is fifteen. That gives you a buffer for a good week without burying the product.
| Slot example | Sells per week | Service interval | Par | Coil capacity |
|---|---|---|---|---|
| Top seller, cola 20oz | 40 | Weekly | 60 | 12 per column, so 5 columns |
| Mid mover, salted chips | 12 | Weekly | 18 | 10, filled to 9 across 2 columns |
| Slow mover, premium bar | 3 | Weekly | 5 | 10, filled to 5 only |
| New item on trial | Unknown | Weekly | 4 | 6, one column |
Two things follow. Par can sit below capacity, and for slow sellers it should. And a product selling forty units a week cannot live in one column of twelve, because it empties in two days.
Turns decide whether a slot earns its space
Turns are the only number in vending machine inventory management that tells you whether a slot earns its space. A slow row is not neutral. It occupies a slot that could hold something that sells, and it will eventually cost you the product outright. The check is mechanical. Any slot that has not sold through a full case in six weeks comes out and something else goes in.
Run it monthly, on paper or in a spreadsheet, and per machine rather than across the fleet. A product that turns in one building and dies in another is telling you about the building, not the product.
The corollary is that a slot you cannot turn is not variety. It is inventory aging in public, and the customer can see it. Sun faded packaging in the bottom row is the most expensive kind of marketing.
Shelf life is where the money leaks
Shelf stable is not shelf eternal, and this is where most vending machine inventory management goes wrong. Every category has its own shrink profile, and the ones that look most appealing on a buying trip are usually the worst offenders.
| Category | Typical shelf life | Shrink as share of category sales |
|---|---|---|
| Chips and salty snacks | 4 to 6 months | 1 to 2% |
| Candy and gum | 9 to 12 months | Under 1% |
| Cookies and crackers | 5 to 8 months | 2 to 3% |
| Pastries and honey buns | 2 to 3 months | 6 to 10% |
| Protein and granola bars | 6 to 9 months | 2 to 3% |
Across a mixed snack and drink machine, a shrink allowance of 2 to 3% of gross is realistic. On a route doing 10,000 USD a month that is 2,400 to 3,600 USD a year, and stales are a large part of it. Above 3%, something structural is wrong, and it is almost never theft.
Here is the part that changes how you stock. Do the arithmetic on one column. A spiral holding twelve units, filled every visit, on a product that sells two a week, holds six weeks of inventory. Most vending snacks have a shelf life measured in a few months from manufacture, and a case bought a month ago is already partway through its date. That slot has almost no margin, and it will produce stales as a matter of physics no matter how careful you are on visit day.
The same coil holding an item that sells eight a week turns in a week and a half and never sees a date code. So the response to a stale pull is not more vigilance. It is a different product, or a smaller fill.
FIFO means unloading the coil
First in, first out is the part of vending machine inventory management that gets faked most often, and the oldest rule in the trade. Dropping new units into the front of a full coil and pushing the old ones deeper is not rotation. It guarantees that the oldest unit in the machine is the one that never sells.
Done properly, on a spiral machine, rotation means pulling the coil forward, taking the remaining stock out, loading new product behind, and putting the older units back in front. It costs seconds per slot and it is the only version that works.
Two habits make it stick. Pull anything within two weeks of its date rather than waiting for the date itself, because a customer judges freshness by taste and not by a printed number. And write down what you pulled and from which slot. Three visits of that log tells you exactly which slot to change, which is information no report will give you.
Restock frequency is a revenue lever
An empty slot is a lost sale you never see. The customer does not wait, does not complain, and does not come back tomorrow for the same item. They buy a soda instead and you never know it happened.
That is why service frequency moves revenue more than a small price change does. Going from every fourteen days to every ten at a strong site commonly adds 8 to 12% to monthly gross with almost no extra labour, because the drive was happening anyway. You are not adding trips so much as shortening them.
Capacity matters too. A snack machine carries 25 to 40 selections with four to six units per spiral, and a fast row at a good site empties in four days. If your top three products cannot survive the interval, the interval is wrong, not the products.
Pick lists and the van
Loading a van by memory is how a driver arrives with eleven cases of the wrong energy drink and no water. Build the pick list at the warehouse, from the current slot counts, and pack each machine into its own tote before the route starts. Pre-kitting turns a forty minute stop into a twenty minute one and removes the most common cause of overloading.
The pick list should also carry the planogram, so the driver is not deciding placement at the site. Placement belongs to whoever reads the sales data, and that is rarely the person holding the tote.
What telemetry can and cannot tell you
Remote monitoring helps, with one caveat. Sales by payment method, fault logs and temperature are reliable because the machine reports them directly. Slot level counts are inferred and drift after a planogram change, a coil jam or a mis-load, so treat them as a hint rather than a count you can order against. What it does well is flag machines that need a visit before the scheduled day. The remote monitoring breakdown covers the fee structures.
If you are setting up a first route and want a second opinion on the planogram before you commit to a product mix, our team can look at the slot map with you through the contact page. The mix question usually resolves faster once the turns data exists, and the product selection guide covers which categories survive a long service interval. For the cleaning and fault side of the same visit, the maintenance checklist pairs with this routine.
Two routes, two outcomes
An operator in Grand Rapids ran eighteen machines on a fixed Tuesday route for four years. The route was efficient and the machines were full, and he could not understand why gross had been flat for three quarters.
He ran a turns check for the first time in the spring. Fourteen slots across the fleet had not sold a full case in six weeks, while nine products that did sell were crammed into single columns and sold out by Thursday. He cut the fourteen, widened the nine to two facings each, and put two trial items in the freed slots. Gross rose about 11% in the following quarter, and warehouse inventory fell about 6%, because he was no longer buying cases for slots that did not move.
The second case is about a single product. An operator in Tampa had honey buns in three facings across four machines, and could not work out why his shrink number was running near 9% of category sales. The pastry had a three month window, the machines were in buildings with no morning crowd, and the facings were filled to capacity on every visit.
He cut it to one facing, filled to half, and stocked it only on the Monday visit. Shrink fell under 3%, and the freed slot went to a breakfast bar that turned every ten days. He had also bought a pallet of an energy drink at 30% off the year before. It was six months of supply for his volume, and half expired on his shelf before reaching a machine. A discount that outlasts the shelf life is not a discount.
A weekly routine that takes fifteen minutes
- Compare the pick list against actual load and note the variance on each machine.
- Check the back unit of every coil, not the front one.
- Pull anything within two weeks of its date and log the slot.
- Flag any slot that has not moved a full case in six weeks.
- Confirm the top three sellers have more than one facing.
If you are deciding whether to add a machine, rebuild the planogram for an existing one first. It is the cheapest test of whether the problem is the site or the stocking, and it costs a spreadsheet and one afternoon. When you are ready to talk through a specific route, reach out with the machine count and the service interval, and we can work from there.
الأسئلة المتداولة
How often should a vending machine be restocked?
Weekly is the default at ordinary volume, twice weekly where top sellers empty in four or five days. The right interval is the one where your fastest product survives until the next visit. Moving from fourteen days to ten often adds 8 to 12% to gross.
What is a par level on a vending machine?
Par is the quantity you fill a slot to on a normal visit, which is not the coil capacity. A common starting rule is one and a half times weekly sales, rounded to what the coil holds. Slow sellers should be filled below capacity so the product turns before its date code.
How do I stop products expiring in a vending machine?
Shorten the fill, not the visit. A twelve unit coil on a product selling two a week holds six weeks of stock, which is most of a shelf life, and that single arithmetic point explains most of vending machine inventory management. Rotate properly by unloading the coil and putting new stock behind old, and pull anything within two weeks of its date rather than waiting.
What is a normal spoilage rate for a vending route?
Two to three percent of gross revenue on a mixed snack and drink route is realistic. Pastries run far higher, often 6 to 10% of their category sales. Above 3% overall, the cause is usually slow slots rather than theft.
Should I give my top seller more than one slot?
Yes. A single column of a fast mover sells out first and then sets the restock date for the whole machine. Two or three facings on the top three products usually lifts gross more than adding a new item, because it protects the sales you already have.
Can I trust telemetry for inventory counts?
Trust it for sales by payment method, faults and temperature. Slot level counts are inferred and drift after a planogram change or a coil jam. Use them to spot a machine that needs an early visit, and count the coil yourself when you are in front of it.
