Purchase Orders and Vendor Management for Traffic Control Companies

Procurement is not the first thing a growing traffic control company fixes, and that is usually correct — dispatch, tickets, and billing all matter more. But it is a reliable second-order leak, and unlike most leaks it is entirely invisible until someone goes looking.

The mechanism is simple. At three crews, the owner knows about every purchase because they made it or were standing there. At twelve, several people can commit money, nobody sees the aggregate, and the company discovers in the annual review that it spent considerably more on cones than anyone believed.

What a purchase order is actually for

Companies often implement purchase orders as paperwork and then resent the paperwork. It helps to be clear about the three controls that justify it.

Approval before commitment. Someone with authority agreed to the spend before it happened. This is the entire point. An approval that occurs when the invoice arrives is not an approval, it is an acknowledgment.

Price agreed before invoice. The number was settled in advance rather than discovered later. Absent a PO, disputes about price are argued from memory against a vendor's paperwork, which is not a strong position.

Three-way match. Ordered, received, invoiced — all three agree. This is what catches short deliveries, duplicate invoices, and the surprisingly common case of being billed for something returned.

Everything else — numbering conventions, approval tiers, forms — is machinery around those three. If a process is not delivering one of them, it is overhead.

A process that fits a company this size

Elaborate procurement kills itself in a field-service business. A workable minimum:

Thresholds, not universal POs. Below some amount, a named person can buy on a company card with a receipt. Above it, a PO. Set the threshold high enough that routine consumables do not generate paperwork and low enough that meaningful spend gets reviewed.

One approver per tier, with a named backup. The most common failure is not lack of control, it is a single approver who is in the field, which pushes people to buy first and paper it later. The backup is what keeps the process alive.

Receiving is recorded by whoever receives. In the yard, at delivery. Not reconstructed from a packing slip found later.

POs reference a job where one applies. Job-specific purchases — a rental for a particular closure, a sign fabricated for one site — should carry the job number so the cost lands where it belongs. Otherwise job costing quietly becomes fiction, and you keep bidding work whose true cost you have never seen.

Emergency purchases have a defined path, not a prohibition. Crews will occasionally need something now. A rule that forbids it produces either delays or quiet rule-breaking. A rule that permits it with same-day notification produces neither.

The vendor categories that matter

Traffic control purchasing splits into a few distinct relationships, and treating them identically is a mistake.

Device manufacturers and distributors. Cones, drums, barricades, sign panels and blanks, sheeting, stands. The largest recurring spend and the one where volume commitments and annual pricing actually move the number. This is also where device condition standards feed procurement directly — a planned replacement cycle derived from your own retirement data is a far better negotiating position than a series of urgent orders. See sign retroreflectivity and device condition standards.

Equipment rental. Arrow boards, message signs, portable signals, attenuator vehicles for peaks. The relationship matters more than the rate here, because availability during your busy period is the thing you are actually buying. A vendor who takes your call in July is worth more than one who is a few percent cheaper in February.

Fleet and vehicle. Trucks, maintenance, tires, fuel. Frequently the second-largest cost category and frequently the least managed, because it is spread across many small transactions.

Fabrication. Custom signs and project-specific panels. Usually lead-time-sensitive rather than price-sensitive, which means the vendor conversation should be about turnaround commitments.

Safety supplies. High-visibility apparel and PPE. Small spend, real consequences — garments degrade and need replacing on a schedule rather than when they fall apart. See high-visibility apparel classes explained, and note that employer obligations sit with OSHA, whose highway work zones material and construction standard at 29 CFR Part 1926, Subpart G are the reference points. Device acceptability itself traces to the Manual on Uniform Traffic Control Devices as adopted by your state, so confirm what your jurisdiction accepts before committing to a large order of anything.

Buy versus rent, honestly

The buy-or-rent question comes up constantly and is usually answered by instinct. The instinct is generally to buy, because ownership feels like capability.

The calculation that matters is utilization. A device used most weeks justifies ownership. A device needed for four peak weeks a year usually does not, once you include storage, maintenance, transport, and the capital tied up.

Two things make companies get this wrong. Utilization is estimated optimistically — everyone remembers the weeks the equipment was busy and forgets the months it sat. And the storage and handling cost of owned equipment is invisible, because it shows up as yard space and labour rather than as a line item.

The honest version requires actual data: how many days per year was this class of equipment deployed and earning? Which is the same record that drives rental billing and device tracking by job site. Companies that have it make this decision well; companies that do not make it by feel and end up with too much of the wrong equipment.

There is a strategic dimension too. Owning the equipment your bids depend on removes a dependency, which matters when a competitor is also your rental supplier — a situation more common in this industry than in most, and worth being deliberate about when bidding municipal contracts against companies you also buy from.

Where the savings actually are

Owners tend to look for savings in unit price. That is the smallest available lever.

Buying less, because the yard record is right. Emergency retail purchases happen because someone thought there were forty cones and there were twenty. This is the single largest source of avoidable spend in most companies, and the fix is an inventory record rather than a negotiation.

Fewer replacements, because devices come home. Devices lost on jobs are replaced at full cost and often not billed to anyone. A per-site deployment record makes loss traceable and chargeable — see charging for damaged and lost devices.

Fewer rentals, because owned equipment was findable. Renting an arrow board while owning one that is sitting on a completed job is a real and common occurrence.

Planned replacement instead of urgent replacement. Scheduled purchasing gets better pricing and better lead times than reactive purchasing, always.

Every one of those depends on knowing what you own and where it is — which means procurement discipline is downstream of inventory discipline, not parallel to it. Traffic OS carries purchase orders, vendors, and multi-yard inventory alongside the jobs the equipment is deployed to, on flat-tier pricing by company size. If your last quarter included an emergency device purchase you are not sure was necessary, book a walkthrough.