The only document that knows the whole job
Walk any small shop floor and you will find the same object sitting next to a part in a tote: a folded, marked-up sheet with a part number at the top and a list of operations down the side. Somebody printed it when the job was released, and it has been travelling with the work ever since. That is the job traveler, and depending on who trained the person who set up your paperwork you might call it a router, a route sheet, a manufacturing traveler, or just the packet.
It is easy to think of it as filing. It is not. On most shop floors the traveler is the single source of truth about a job, and often the only one. The scheduling board says the job exists. The quote says what you promised. The traveler is the only thing that knows what has actually happened to this part, in what order, by whom, and what is supposed to happen next.
Which makes it worth asking what is on yours. Most travelers are inherited. Someone made the template years ago, it worked well enough, and it has been photocopied forward ever since without anyone reviewing what it captures. The fields that are on it get filled in. The fields that are not on it do not get thought about, and that is where the cost hides.
If you want a working starting point rather than a critique of the one you have, there is a free job traveler template that lays out the standard fields and prints cleanly. This article is about the ones that usually go missing.
The fields nearly every traveler has
Start with the baseline, because the omissions only make sense against it. A functional traveler carries:
- Job number and part number. The identifiers that tie this piece of paper to the order, the quote, and the drawing.
- Customer and due date. So anyone holding the sheet knows who is waiting and how long they have been.
- Quantity. How many were released.
- Material. What it is made from, and ideally where that material came from.
- The operation list. Each step in sequence, with the work center it runs on. This is the spine of the document.
- Estimated time per operation. Setup and run, so the operator knows what they are working against.
- Sign-off space. Somewhere for an operator to initial and date each completed step.
Most shops have all of that. If yours does, the template is doing its basic job. Now the interesting part.
What most shops leave off
Actual time, right next to estimated
Almost every traveler has a column for the planned hours. Very few have a column beside it for what the operation actually took.
That single missing column is why shops quote from the same optimistic numbers year after year. The estimate goes out on the traveler, the operator runs the job, the part ships, and the real number evaporates. Nothing ever contradicts the estimate, so the estimate never improves.
Adding one narrow column, planned beside actual, turns every job into a free data point. After a few weeks you start seeing patterns you could not see before. That one alloy always runs long. The deep pocket parts are always underquoted. The estimate is fine until setup, and setup is where the time leaks. This is the mechanism behind most late jobs, and the traveler is where you catch it, because the traveler is the only document standing next to the machine while it happens.
It is also the input your shop rate depends on. You cannot know your real billable hours if nothing on the floor ever records how long the work took.
Setup time kept separate from run time
Related, and worth its own field. Many travelers lump the whole operation into one number.
Setup and run behave completely differently. Setup is mostly fixed regardless of quantity. Run scales with the piece count. Blend them and you lose the ability to quote a different batch size intelligently, because you have no idea which part of the number moves with quantity and which does not.
Splitting them also makes your changeover cost visible for the first time. Most small shops are carrying 10 to 30 percent more capacity than they think, sitting in setup time nobody has ever measured, and you cannot go looking for it until the traveler stops hiding it inside a combined figure.
Material heat or lot number
If you do anything near aerospace, medical, or defense, this is not optional and you already know it. If you do not, it is still the field that saves you the day a customer calls about a part you shipped fourteen months ago.
Traceability means being able to walk backward from a finished part to the specific material it came from. The traveler is where that link gets made, because it is the one document present at both ends of the job. Write the heat or lot number on it at material issue and the chain holds. Leave it off and the chain breaks at the first step, permanently, and no amount of searching later will rebuild it.
The drawing revision
Not the drawing number, the revision.
This one bites in a specific and expensive way. A customer sends a revised print. The office files it. The job on the floor is running to the traveler that was printed three weeks ago, which points at a drawing number that has not changed, because only the revision changed. The parts come out perfectly to the wrong version.
Putting Rev on the traveler next to the drawing number costs nothing and gives everyone a chance to catch the mismatch. It is also exactly what configuration control under AS9100 is asking for, if you are heading that direction.
The operations that do not happen on a machine
Traveler operation lists have a habit of listing only the machining steps. Op 10 mill, op 20 lathe, op 30 mill. Then the part needs deburr, and inspection, and cleaning, and packing, and none of those are on the sheet.
They still take time. They still tie up a person and a bench. And because they are not on the traveler, they are not in anyone’s estimate and not on the schedule, which is why jobs mysteriously take longer than the sum of their machining operations. Work that is invisible on paper is still visible in the calendar.
If a step consumes time or has to happen before the next one can, it belongs on the list, even if no spindle turns.
Outside processing, with dates out and back
This is the biggest single omission, and it is the one that costs the most.
The moment a job leaves for heat treat, plating, anodizing, or grinding, it leaves your floor. It also leaves everyone’s attention, because attention on a shop floor follows what you can see. Meanwhile the clock the customer is watching keeps running.
An outside operation needs its own line on the traveler with three things: the vendor, the date it went out, and the promised date back. Without those, “at heat treat” is a status with no expiry, and jobs can sit in it for two weeks past the promise before anyone notices. Shops that start tracking outside process dates properly are almost always surprised by what they find, because it was the one part of their lead time nobody was measuring.
Quantity at each step, not just at the start
Travelers usually record the quantity released. Fewer record the quantity that came out of each operation.
Parts get scrapped. If the traveler says 50 and op 40 is running 47, you want that discovered at op 40, not at final inspection when there is no time left to make three more. Recording quantity at each sign-off turns a nasty end-of-job surprise into a manageable mid-job one, and it tells you where in the process you are actually losing parts, which is the first step to losing fewer.
The gotcha from last time
The most valuable thing on a well-used traveler is often handwritten in the margin. Soft jaws in the second drawer. Do not trust the print on the bore depth, customer confirmed the deeper one. Runs better at 80 percent feed.
That knowledge exists in your shop right now, in one or two people’s heads. It walks out the door at retirement, and it is unavailable on the night shift. A traveler with a deliberate notes field, and a habit of reading those notes forward into the next run of the same part, is the cheapest knowledge capture a small shop can do. Most shops have the margin notes. Very few have the habit of carrying them forward.
What a traveler cannot do
Here is the limit worth being clear about, because a lot of shops try to make the traveler carry more than it can.
The traveler tells you the route. It cannot tell you the timing. It knows this job needs the mill and then the lathe, but it has no idea that four other jobs also want the mill on Thursday. It is a document about one job, and scheduling is a question about all of them at once.
That is why shops with excellent travelers still miss dates. Every individual job is documented perfectly and nothing coordinates them. The traveler is the record. A scheduling board is the plan. You need both, and neither substitutes for the other.
The second limit is that a paper traveler exists in exactly one place. Whoever is holding it knows the status of that job. Nobody else does. When the office wants to answer a customer asking where their order is, someone walks the floor looking for a piece of paper. That works at four jobs and becomes a real cost at forty, which is the same arithmetic behind tracking jobs across multiple machines.
When paper stops being enough
Paper travelers are not primitive. They are genuinely good at what they do: they are readable in a noisy shop with dirty hands, they need no battery, and everyone already knows how to use one.
The signals that a shop has outgrown them are specific:
The traveler is the only copy. When a lost packet means a lost job history, one coffee spill is an incident. Any document you cannot afford to lose should not exist in exactly one place.
The office cannot answer status questions. If “where is job 4412” requires a walk rather than a look, you are paying for that walk several times a day.
Nobody is reading the actuals. Operators write times on the traveler, the traveler gets filed in a box, and no one ever opens the box. That is data collection with no feedback loop, which is all cost and no benefit. This is the same trap as running the shop on spreadsheets: the recording happens, the learning does not.
You need history. A customer reorders a part you ran two years ago and you want to know what it really took. If that answer lives in a filing box, it is technically available and practically not.
The move that works is rarely a full switch. Replacing every paper traveler at once in a high-mix shop is disruptive and tends to fail, because you are changing the habits of everyone on the floor during a week when you also have parts to ship. The path that sticks is to keep the paper where it is useful, at the machine, and mirror the job’s status somewhere the office can see it. The operator still initials the sheet. The status just stops being trapped on it.
Start with the sheet you have
You do not need a new system to fix most of this. You need to look hard at the template you are already printing and ask what it is not asking for.
If you add nothing else, add two columns for actual time and one line per outside process with dates. Those two changes cost you a revised template and they close the two largest blind spots most small shops have: they do not know what their work really takes, and they do not know how long their jobs sit at vendors.
Everything else on the list above is worth doing and can wait. Those two pay for themselves on the first job.
For CNC shops especially, where the whole business is selling accurately estimated spindle time, a traveler that captures what actually happened is not paperwork. It is the only feedback your estimating has, and right now it is probably sitting in a box in the office, half-filled-in, waiting for someone to read it.