A driver pulls into a parts plant at ten past nine, walks in empty handed, opens the machine, sees what is low, walks back to the truck, digs through four cases for the right flavor, walks back in, fills eight columns, walks back out for the drinks, walks back in again. Forty minutes for a stop that should take twelve. Two of those columns were already full.<\/p>\n
Nobody taught him to work that way. The truck is stocked as a rolling store, a bit of everything in case something sells, and the driver shops from it at the machine. Vending machine pre-kitting is the fix, and it is less about software than about where the picking happens. Move the decision to the warehouse the night before, pack each machine’s fill into its own labeled tote, and the stop becomes a single trip with a written list.<\/p>\n
\nKey takeaways<\/strong><\/p>\n<\/blockquote>\n
><\/p>\n
\n\n
- A stop is not eight minutes at the glass. It is drive time, dwell time and vehicle cost, and it runs about 19 dollars before you sell anything.<\/li>\n
- Pre-kitting means picking each machine’s fill in the warehouse from a par minus on hand calculation, not shopping from the truck at the door.<\/li>\n
- Two pick lists are needed. One by machine for the driver, one by product for whoever picks.<\/li>\n
- The numbers that prove it are pick accuracy, returns per kit, sold-outs on arrival, expired units pulled and minutes on site.<\/li>\n
- Roll it out on one stable route first, and let the driver correct the estimate at the machine.<\/li>\n<\/ul>\n<\/blockquote>\n
The double trip you are paying for<\/h2>\n
Start with the cost of a stop, because that number decides whether pre-kitting is worth the effort.<\/p>\n
Take a driver at 24 dollars an hour, which is 40 cents a minute, and a vehicle at 65 cents a mile. Twenty minutes of driving plus fifteen on site is 35 minutes, or 14 dollars of labor. Eight miles round trip is 5.20. Total stop cost, 19.20 dollars.<\/p>\n
Now divide by gross margin. At 55 percent, a stop needs about 35 dollars of sales just to break even. Below that, the trip loses money whether the machine is empty or full.<\/p>\n
Every minute you take out of dwell time is worth 40 cents, and every stop you can add to a day is worth the contribution on a machine that was previously skipped. A driver doing twelve stops and shaving five minutes off each one gets an hour back. That hour is either a thirteenth stop or a truck that reaches the depot before the counting room closes.<\/p>\n
The double trip is where the minutes hide. Walking back to the truck for a missing flavor costs three to five minutes, and it happens on most stops when the truck is a rolling store. Multiply by twelve stops and five days and you are looking at three hours a week of walking, which is 72 dollars, which is four stops.<\/p>\n
What a kit actually contains<\/h2>\n
In vending machine pre-kitting, a kit is the exact fill for one machine, packed in one tote, labeled with the machine number and the route position.<\/p>\n
The arithmetic behind it is simple and unforgiving. For every selection, take the par, subtract the estimated on hand, and round the result to what the coil can physically take. Par is either the full coil capacity or the level you have chosen to fill to, and those are not always the same number. A slow seller with two units left and a par of ten may only deserve four.<\/p>\n
Three rules keep the arithmetic honest.<\/p>\n
\n
- Round down to what you want in the machine, not up to what fits.<\/strong> A kit that tops every column to capacity is a kit that manufactures stale stock.<\/li>\n
- Keep one exception list.<\/strong> New items, seasonal pushes and a manager’s request all break the formula. Write them down rather than letting the driver improvise.<\/li>\n
- Do not kit what the machine already has.<\/strong> The estimate is the whole point. If the estimate is wrong, the count at the machine fixes it, and the correction is what sharpens the next kit.<\/li>\n<\/ul>\n
Then convert the lines into whole cases, because that is how the warehouse is shelved. The remainder from an opened case becomes tomorrow’s first pick, which is why a warehouse that kits consistently ends up with less open inventory than one that does not.<\/p>\n
Two pick lists, not one<\/h2>\n
The most common failure in a first vending machine pre-kitting setup is producing a single document and handing it to the wrong person.<\/p>\n
Whoever picks needs the list by product. Twelve machines on tomorrow’s route, each wanting four bags of the same chips, means 48 bags pulled from one shelf position, not twelve separate walks to the same shelf. That view is what makes picking fast.<\/p>\n
The driver needs the list by machine. Standing at the glass, he needs to know that this column gets six and that one gets four, in the order he will fill them.<\/p>\n
Both views come from the same planogram record<\/a>. Produce one without the other and the system collapses into either slow picking or an unreadable clipboard at the door.<\/p>\n
A scanner closes the loop. Scanning each unit as it goes into the tote catches the wrong size or the wrong flavor in the warehouse, where the fix costs ten seconds, rather than at the machine, where the fix costs a return trip and a refund.<\/p>\n
Load the truck backwards<\/h2>\n
In vending machine pre-kitting, totes go into the vehicle in reverse route order. The first stop of the day goes in last, nearest the door. The last stop goes in first, at the back.<\/p>\n
This sounds like a small thing until you watch a driver unloading half a truck at six in the morning to reach the tote for stop one. Reverse loading is free, it takes no extra labor, and it saves a few minutes every morning for the life of the route.<\/p>\n
Two warehouse habits pay for themselves quickly.<\/p>\n
Take the stock out of inventory when the kit is marked complete, not when the machine is filled.<\/strong> Otherwise the warehouse count is a day wrong forever, and every pick list built on it inherits the error.<\/p>\n
Keep a staging lane per route.<\/strong> Totes for route three sit together on one shelf in stop order, checked once before the truck leaves. A missing tote found at the depot costs a phone call. Found at the machine, it costs a stop.<\/p>\n
What changes at the machine<\/h2>\n
The visit gets short and repeatable, which is the point of the whole exercise.<\/p>\n
The driver arrives with one tote and a list of add quantities per selection already filled in. He confirms or corrects the count he finds, adds the kit, pulls anything expired or dead with a reason, collects the cash, pulls the audit file if the machine produces one, and closes up.<\/p>\n
A count only matters where the estimate was wrong. Those corrections are the raw material for the next estimate, which is why entering them at the machine rather than after the route matters. Counts typed from memory at six in the evening are how a good system drifts into a bad one.<\/p>\n
The other habit is recording a skipped stop as a visit with a reason. Location closed. Machine full, eleven days. Could not park. A skip that leaves no trace makes the machine look serviced when it was not, and the next estimate is built on a fiction.<\/p>\n
The warehouse side of the change<\/h2>\n
Pre-kitting adds work to the warehouse and takes work off the road. That trade is only worth it if the warehouse work is organized.<\/p>\n
Slot the fastest moving items nearest the picking station. Keep a fixed position for every SKU and do not move it for a promotion, because pickers learn positions rather than names. Stage empty totes in route order, and put the printer or tablet where the picker stands rather than in an office.<\/p>\n
The case study that gets quoted most often is Maple City Dispensing, a family business in Goshen, Indiana that has run since 1984. It operates close to a thousand machines across 22 markets on what were eight routes, one of them part time. Before pre-kitting, trucks were loaded with whatever might be needed and drivers made product decisions on the fly. After moving to a warehouse picking system, picking time roughly halved, the team went from picking one machine at a time on paper to picking three at once, routes came down from eight to six, and drivers started finishing about an hour earlier each day.<\/p>\n
Read that sequence carefully. Picking did not get faster because the software was clever. It got faster because the work moved to a place where one person picks for three machines at once instead of walking to the same shelf three times.<\/p>\n
Five numbers that say it is working<\/h2>\n
Track these five numbers weekly per route. They cost nothing to collect and they tell you whether vending machine pre-kitting has actually happened.<\/p>\n
\n\n
\n \nMetric<\/th>\n How to read it<\/th>\n What a bad number means<\/th>\n<\/tr>\n<\/thead>\n \n Pick accuracy<\/td>\n Correct lines divided by total lines<\/td>\n Warehouse layout or verification is failing, not the picker<\/td>\n<\/tr>\n \n Returns per kit<\/td>\n Units returned divided by units picked<\/td>\n Par levels are set too high<\/td>\n<\/tr>\n \n Sold-outs on arrival<\/td>\n Empty columns found at each visit<\/td>\n Interval too long or par too low<\/td>\n<\/tr>\n \n Expired units pulled<\/td>\n Items past date removed per visit<\/td>\n Kits are too big for the sales rate<\/td>\n<\/tr>\n \n Minutes on site<\/td>\n Arrival to departure per machine<\/td>\n Driver is still re-picking at the glass<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n The last one is the honest test. If minutes on site did not fall, pre-kitting has not happened, no matter how good the pick list looks.<\/p>\n
Where vending machine pre-kitting breaks<\/h2>\n
Stale par levels.<\/strong> A kit is only as good as the par behind it. Machines that got a new planogram in March and a par update in November produce kits for a machine that no longer exists.<\/p>\n
Planogram drift.<\/strong> If the warehouse thinks column B4 holds pretzels and the machine holds peanuts, every kit for that machine is wrong by one line, every visit, until someone updates the record.<\/p>\n
Fresh food.<\/strong> Short dated items get picked on the morning of delivery, not the night before, and never sit in a tote in a warm warehouse.<\/p>\n
No data.<\/strong> Without telemetry or audit files the estimate has to come from counts. That works if the driver counts every machine on the next two visits and the counts get entered. Two visits build a sales rate good enough to kit from.<\/p>\n
Kitting for machines that should not be visited.<\/strong> Pre-kitting makes a route efficient. It does not rescue a bad one. A machine that sells 20 dollars a week is still a candidate for relocation, not for a better tote.<\/p>\n
Rolling it out without wrecking a route<\/h2>\n
Do not convert a whole operation to vending machine pre-kitting in a week. Convert one route with stable sales and machines that already report data.<\/p>\n
Start with truck kitting, which is the same discipline with a lower bar. The driver packs one tote per machine from vehicle stock before the first stop. No warehouse staff, no new software, and the habit of working from a list instead of from a shelf gets established.<\/p>\n
Then move picking into the warehouse for that route. Involve the driver in the first two weeks, because he knows which machines sell out first and which columns are a waste of space. A pick list that ignores what the driver knows will be overridden at the machine, and once a driver starts overriding it, the system is dead.<\/p>\n
Add routes one at a time. Expect the first month to feel slower, because a picker who has never worked from a list is slower than a driver who has shopped from a truck for nine years.<\/p>\n
What to do this week<\/h2>\n
Pick your steadiest route and export one day of sales by column. Build the fill list by hand for six machines, pack six totes, and label each one with the machine number and the stop order. Load them in reverse and run the day.<\/p>\n
Then answer three questions that night. Did minutes on site fall. Did any kit come back with units still in it. Did the driver have to walk back to the truck for anything.<\/p>\n
If you are setting up a warehouse for the first time, the warehouse layout guide<\/a> covers shelving and slotting, and the restocking frequency guide<\/a> covers how the visit interval feeds the par calculation. Both of those decisions come before the tote.<\/p>\n