The most expensive place a job can sit
Every field-service operation has a graveyard it does not like to look at: the jobs that got diagnosed, quoted, sometimes half-repaired, and then stalled because a part had to be ordered. The tech did good work, the customer wants the fix, the money is real — and then the job goes quiet, waiting on a compressor or a control board or a pump seal that is somewhere between a distributor's shelf and your customer's door. These jobs are uniquely dangerous precisely because nothing is obviously wrong. Nobody dropped the ball in a way you can see. The job is just... paused. And paused jobs, left unmanaged, rot.
They rot in specific, costly ways. The part arrives and sits in the shop because no one connected the delivery back to the waiting job. The customer, hearing nothing, assumes they have been forgotten and calls a competitor. The tech who did the diagnosis moves on to two weeks of other work and loses the context, so the return visit starts half-cold. The special-order part that was paid for gets misplaced, or worse, quietly installed on a different job. None of this is exotic — it is the ordinary entropy of a job that left the normal flow and entered a waiting state that most operations have no real system for. The parts-hold problem is not a parts problem. It is a tracking problem wearing a parts problem's clothes.
Name the state so the system can see it
The first and most important move costs nothing and fixes half the problem: give "waiting on parts" a real, distinct status, and use it. A job blocked on a special-order part is not the same as a job in progress and it is not the same as a job scheduled — it is its own thing, defined by the fact that it cannot advance until a physical object arrives. When that state is nameless, the job hides inside "in progress" and becomes one of the jobs stuck in progress that you only discover weeks later when you go looking for why revenue leaked. When the state has a name, the job becomes a queryable, visible thing: you can pull every job held for parts, see how long each has waited, and work the list.
That visibility is the whole point. A parts-hold status turns an invisible graveyard into a managed queue. The dispatcher can look at it every morning; the office can see which holds are aging past their expected arrival; nobody has to remember a waiting job because the system remembers it for them. Hosting Field's 7-state job model exists exactly so that a job in an unusual condition — waiting on a part, paused, blocked — has somewhere honest to sit instead of being crammed into a status that lies about what is really happening. The discipline is simply to use the state the moment a part gets ordered, not to leave the job marooned in "in progress" and hope someone circles back.
Tie the part to the job, both directions
A parts-hold status tells you a job is waiting; it does not, by itself, tell you what it is waiting for or when the wait ends. That connection has to be made explicit, and it has to run in both directions. On the job side, the record should carry what was ordered, from whom, when, and the expected arrival — so anyone opening the job sees not just "held for parts" but "held for a specific pump, ordered Tuesday, expected Friday." That single line converts a vague pause into a plan with a date, and a date is something you can manage against.
The other direction is the one operations forget, and it is where jobs die. When the part physically arrives, the delivery has to find its way back to the job that was waiting for it. If receiving a box and reviving a stalled job are two disconnected events, the box will sit on a shelf while the job sits in the queue, each unaware of the other. The link between parts procurement and the purchase order that placed the order is what closes that loop: the PO knows which job it was raised for, so a received PO can point straight back at the job it unblocks. Get that connection right and "the part came in" automatically becomes "this job is ready to schedule" instead of a fact that lives only in the receiving clerk's head.
Keep the special-order part distinct from your rolling stock, too. A part ordered for a specific job is not the same as the general truck stock and inventory a tech carries — it is committed, spoken for, and should never be quietly consumed by a different call because someone needed one in a hurry. Reserving the part against its job protects the customer who is waiting from having their fix cannibalized by the next emergency. This is also the practical counterweight to stocking the truck to avoid second trips: truck stock prevents the parts-hold from happening at all for common items, and disciplined reservation protects the holds you could not prevent.
The customer is waiting too — say something
While the part is in transit, the customer is living inside an information vacuum, and that vacuum is where goodwill dies. From the customer's chair, a job held for parts is indistinguishable from a job forgotten unless you tell them otherwise. They do not know a distributor was out of stock or a manufacturer put the board on backorder; they know only that a tech came, said they would be back, and then silence. Silence reads as neglect even when the truth is a supply chain you do not control. The fix is almost insultingly simple and almost universally skipped: proactively tell them the part is on order, roughly when it is expected, and that you will schedule the return the moment it lands.
Set the expectation honestly, including the uncertainty. "The control board is on order; distributors are quoting three to five business days, and I will call you the day it arrives to book the return" is a promise you can keep, and it reframes the wait from abandonment into a managed process the customer is being carried through. This is the same keeping-customers-informed discipline you apply everywhere else in the job lifecycle, just aimed at the one moment customers most often feel dropped. If the expected date slips — and with special orders it will sometimes slip — tell them before the old date passes, not after. A proactive "the part got delayed to next week" preserves trust; a missed date you never mentioned burns it. Two-way messaging makes this cheap, and the same customer communication habits that carry a normal visit carry a parts-hold wait if you simply remember to use them.
Close the loop the day the part lands
The whole parts-hold workflow exists to make one moment effortless: the return trip. When the part arrives, the goal is that scheduling the finish is a fast, obvious next step rather than an archaeology project. Because you named the state, linked the part to the job, and reserved it, the person receiving the delivery can move the job straight from "held for parts" back into the schedulable pool — and the dispatcher slotting it in has the original diagnosis, the customer's context, and the reserved part all attached. The return visit starts warm.
Schedule the return deliberately, not whenever a gap happens to appear. A customer who has already waited on a part has spent their patience; the finish should be prioritized, not treated as filler for a slow afternoon three days later. Where the return is more involved than a quick swap, treat it with the same rigor you would give any multi-visit job — the first visit diagnosed and ordered, the second completes, and the two are explicitly halves of one job rather than two disconnected calls that happen to share an address. Keeping them linked also protects your numbers: a parts-hold that is invisibly split into two jobs quietly wrecks your first-time-fix-rate reporting, making it look like you failed to fix on the first visit when the truth is you were waiting on a part nobody stocks. Measured honestly, the parts-hold is a known, managed detour — not a failed job.
Learn from the holds you keep having
A parts-hold workflow is a safety net, but a net you keep falling into is telling you something. The jobs that repeatedly stall waiting on the same categories of parts are data, and if you read them you can prevent tomorrow's holds instead of just surviving them. When a particular part shows up again and again on the held-for-parts list, that is a candidate to move onto the truck as regular stock — the recurring hold is the business case for carrying it, quantified. When a specific distributor is consistently the source of the longest waits, that is a supplier conversation or a second-source decision, not a fact of nature to keep absorbing. And when holds cluster around a certain kind of repair, it may be a signal to change how you diagnose or quote that work up front so the part is on the truck or on order before the tech is standing in the customer's basement.
The point is that the held-for-parts queue is not just a to-do list; watched over time it is a map of where your parts availability fails, and every recurring entry is a chance to convert a reactive scramble into a stocked part or a better supplier. Warranty parts deserve their own lane inside this same discipline — a defective component going back to a manufacturer under supplier warranty is a parts-hold with a claim attached, and it needs the same visibility so the replacement's arrival, and the credit, both get tracked back to the job.
The honest boundary
Hosting Field will give the waiting job a real place to sit, tie it to the part that was ordered and the purchase order that ordered it, keep the customer informed through the wait, and make the return trip a fast next step instead of a cold restart — and it will keep the held-for-parts queue in front of you so nothing rots in silence. What it will not do is make the distributor ship faster, decide which recurring parts belong on your trucks, or call the customer for you. The software's job is to make sure a stalled job is never a forgotten job: named, visible, linked, and reserved, so the part and the job find each other the day the box arrives. Whether the wait feels like neglect or like a managed process — that is decided by whether you actually work the queue the tool keeps for you, order early, communicate before the dates slip, and treat the return trip as the priority it is. Do that, and "held for parts" stops being where jobs go to die and becomes just another state a job passes through on its way to done.