Do You Make These Procurement Mistakes?
Introduction
The controller closes the quarter and pulls up the P&L. Revenue is flat. Gross margin is down 320 basis points. Nothing big happened. No lost customer, no bad job, no equipment failure. Materials cost more. Freight cost more. Two suppliers raised prices. A handful of expedites landed in March. On-time delivery slipped into the low 80s and the cost of catching up ate the quarter.
Most of the damage is procurement-shaped. Nobody in the shop owns that word.
In almost every small-to-mid-sized US manufacturer, the title "procurement manager" does not exist. There is a buyer, a shop foreman who calls suppliers, and an owner who signs the big POs. The work that should be called procurement is spread across three people and two spreadsheets. The bill for that arrangement shows up in the quarterly P&L, in inventory that does not turn, and in the multiple the business earns on the day it sells.
This article lays out what procurement actually is, why it matters now, and the ten recurring mistakes that hurt job shops, contract manufacturers, precision machining outfits, and electronics assembly shops. Most are fixable with process, not software.
What procurement actually is
Procurement is the strategic process of acquiring external goods and services a business needs to operate. In manufacturing, procurement is responsible for raw materials, components, consumables, MRO supplies, tooling, and the services that keep production running. It owns cost, risk, supplier relationships, and the terms on which materials enter the building.
Three terms get used interchangeably and should not be:
- Purchasing is the transaction. A buyer issues a purchase order, receives the goods, and pays the invoice. Purchasing is a subset of procurement.
- Sourcing is the supplier-selection step. Identify, qualify, and award.
- Supply chain management is the end-to-end flow of materials and information, from supplier's supplier to customer's customer. Procurement is the upstream half.
The procurement lifecycle, in the language most large shops use, covers six stages:
- Demand identification. Engineering, operations, or a customer commitment creates a need. Specs, quantities, and timing are defined.
- Sourcing. Qualified suppliers are identified and evaluated on price, capability, quality, capacity, financial stability, and risk. An RFQ or RFP is run where the spend warrants it.
- Contracting. Terms are agreed. Long-term agreements, MSAs, blanket orders, and pricing formulas replace one-off POs where volume justifies it.
- Purchase order. Approved requisitions become POs. Pricing, delivery, and terms come from the contract or the quote.
- Receipt and invoice matching. Goods arrive. The receiving document, PO, and invoice are reconciled before payment clears.
- Supplier performance management. Quality, on-time delivery, responsiveness, and cost behavior are measured. The data feeds the next sourcing decision.
If your shop is only doing steps 4 and 5, that is not a procurement function. That is a purchasing clerk. Most of the mistakes below are symptoms of the missing first three and last one.
Why this matters right now
The 2025-2026 environment is the hardest procurement climate US manufacturers have faced in a decade. A few data points make the pressure concrete.
The ISM Manufacturing PMI Prices Index jumped to 78.3 in March 2026, the highest reading since June 2022. Supplier deliveries slowed for the fourth straight month. A Supreme Court ruling in early 2026 struck down the IEEPA tariffs, which the administration is replacing with Section 301, Section 232, and Section 122 investigations covering steel, semiconductors, autos, chemicals, machinery, and 16 major trading partners. Tariff policy is no longer a one-time cost. It is a monthly moving variable that procurement has to price into every quote and every contract.
On top of that, 80% of organizations hit at least one supply chain disruption in the past year. 85% of disruptions happen in the lower tiers of the supply chain, where most shops have no visibility. Factory fires and floods were the top disruption categories in Q3 2025. A single burned warehouse at a sole-source supplier can park a production line for weeks.
Any one of these headwinds makes procurement discipline more valuable. All of them at once make the mistakes below expensive in ways they were not five years ago.
The ten procurement mistakes
1. Running procurement as purchasing
The classic small-shop pattern: a PO goes out when engineering or the shop floor says they need something. The buyer gets a quote, picks the lowest number, cuts the PO, and files it. No RFQ history, no supplier capability file, no negotiated contract, no performance data. When the invoice arrives it gets matched, sort of, and paid.
That is purchasing. It is a transactional function executing other people's decisions. Real procurement owns the supplier base, the contracts, and the decision about who gets what work, at what price, under what terms, with what consequences if they miss.
Deloitte's 2025 Global CPO Survey found that 57% of chief procurement officers identify siloed ways of working as the single biggest barrier to value delivery, followed by competing priorities (46%) and technology capability gaps (40%). When procurement is not a function, the savings and supplier accountability it would deliver both stay with the supplier.
2. Letting engineering write the spec alone
Specifications are where money is designed in or out of a part, and engineering writes them before procurement is in the room. The results show up everywhere. An aluminum part called out in stainless because the designer reused a template. A finish spec that costs 3x because a decade-old internal standard is tighter than the industry standard. A proprietary fastener where a COTS equivalent exists at a quarter of the price.
The numbers on this are brutal. When internal engineering standards replace industry standards, companies pay 20% to 100% more than peers who stuck with the standard. In one documented case, consolidating over-specified valves to industry-standard equivalents unlocked $25M in annual procurement savings at a single process manufacturer. These are not exotic examples. They are what happens when procurement is not part of design reviews.
Fix: procurement sits in the DFM/DFA review. Every non-standard call-out has to be defended against a should-cost number, not just an engineer's preference. Clean BOM data is what makes this conversation possible at all.
3. Single-sourcing without a qualified backup
A sole source is a supplier who is the only one capable of providing the part. A single source is a supplier chosen out of several options for commercial reasons. The two get confused, and the second one turns into the first one when nobody maintains a qualified alternate.
Resilinc's Q3 2025 data logged 253 factory-fire disruptions and 169 flood events globally, with flood events nearly doubling year-over-year. Each one is a coin flip on whether a critical tier-two supplier was in the building. Manufacturers who discover a sole source during an incident rather than before it typically lose weeks of production and months of customer credibility.
Dual sourcing carries a real cost. Splitting volume kills bulk pricing, qualification adds engineering and quality work, and managing two suppliers is more expensive than managing one. The right answer is segmented: dual where the part is high-risk and volume supports two viable sources, single where scale and a real partnership outweigh incremental risk. The wrong answer is assuming your single source is a choice when the rest of the market has already consolidated or exited.
4. Flying blind without a supplier scorecard
Ask a small-shop buyer which supplier is best and the answer comes in stories. "The guys in Ohio never miss. The Chicago shop gave us a great price last year but their last shipment was short. The coating vendor is a pain but we have no alternative." Stories are not a scorecard.
A supplier scorecard is five to ten KPIs, measured the same way every month, per supplier. On-time delivery, on-time-in-full, first-pass yield, PPM defects, lead-time variance, response time on quote requests, and invoice accuracy are the core set. Weights reflect what matters for that category.
OTIF (on-time-in-full) benchmarks for suppliers selling to US manufacturers:
| Industry | Acceptable OTIF | Strong OTIF |
|---|---|---|
| Automotive & aerospace | 94% | 97%+ |
| Industrial equipment | 88% | 93%+ |
| Electronics and high-tech | 90% | 95%+ |
| General manufacturing | 85% | 95%+ |
Shops that do not measure supplier OTIF are usually surprised when they finally do. A supplier who felt "pretty reliable" turns out to be hitting 78%. The real cost of that 78% is not in the late shipments, it is in the inventory buffer, the expedites, the safety stock, and the schedule disruption that was quietly being absorbed as "how things are." Published research on lead-time variance finds that reducing variability typically moves inventory more than reducing the average lead time itself.
5. Buying on unit price instead of total cost of ownership
Two quotes come back. Supplier A is $4.20 per unit. Supplier B is $4.65. Most small-shop buyers pick A and move on.
Total cost of ownership (TCO) adds up everything that happens after the PO. Inbound freight, duty, receiving labor, inspection, rework on defects, inventory carrying cost for whatever MOQ the supplier forces, rush-order premiums, and the cost of expediting around the supplier's variability. Research from the Supply Chain Management Review and others has found that TCO-led procurement unlocks up to 30% cost savings versus unit-price buying. Manufacturing Institute data in the automotive sector cites 25% TCO reductions as a realistic ceiling.
The unit-price buyer is one bad batch away from being wrong by ten times the headline savings. The TCO buyer selects on delivered landed economics. Same suppliers, same quotes, different award, different year-end gross margin.
6. Tolerating maverick and tail spend
Maverick spend is buying that bypasses the procurement process. A shop foreman orders MRO supplies off an Amazon Business account. A project manager books a freight lane outside the preferred carrier. An engineering manager signs up for a SaaS tool on a credit card. No RFQ, no contract, no approved-vendor check, no three-way match.
The Hackett Group's 2025 research pegs maverick spend at 20% to 30% of indirect spend leakage in typical manufacturers, and companies lose 5% to 16% of targeted savings to maverick buying in any given year. Top quartile performers lose 60% less to maverick behavior than their peers.
Tail spend (the long tail of small, infrequent, low-dollar purchases) is worse. Over 50% of companies report tail spend above 10% of total spend, while only 4% actively manage most of it. 48% of procurement leaders in the 2025 Hackett Tail Spend Management Study said tail spend had become a materially higher priority.
Fix: a short, obvious PO process with three approval tiers and one approved-vendor catalog covers 90% of this. The remaining 10% gets a quarterly spend review.
7. Treating expediting as a strategy
When the schedule slips, somebody in the shop picks up the phone and air-freights parts in. It feels heroic the first time. By the third month in a row it is the operating model.
Expedited freight costs 2x to 3x standard service. Air freight runs 5x to 10x. Those numbers do not count the supplier's overtime premium, the schedule disruption at the supplier that flows back to you next week, or the gross-margin damage on the order that triggered the rush. Arena Solutions and Life Cycle Engineering both document that sustained expediting is a sign of a broken planning-to-procurement link, not a sign of a responsive shop.
Fix: capture every expedite with a reason code (customer pull-in, supplier late, internal planning miss, quality reject, engineering change, capacity constraint). Review the codes monthly. 80% of expediting is typically concentrated in 20% of the reason codes. Those are procurement problems, not freight problems.
8. Negotiating without a should-cost baseline
The supplier quotes $7.40. The buyer asks for "a better number." The supplier comes back at $7.10. Everybody claims a win.
A should-cost model (sometimes called clean-sheet costing) builds the price of a part from the ground up: raw material mass and cost, cycle time and machine rate, labor minutes, tooling amortization, yield loss, overhead, and margin. The number the model produces is what the part ought to cost. Anything above that is negotiable, and the conversation has numbers in it.
Published case studies routinely report 10% to 36% savings on category spend after introducing should-cost modeling, and one documented example landed $480,000 in annual savings on a single machined component from a two-point negotiation where the supplier kept a healthy margin. The point is not to squeeze suppliers into losses. It is to stop negotiating from the supplier's spreadsheet.
Small shops often believe should-cost requires expensive tooling. It does not. A capable buyer with a spreadsheet, a machine-rate table, and real knowledge of the process can build a defensible should-cost for the top 20 parts that make up 80% of spend.
9. Skipping or faking the three-way match
The three-way match reconciles three documents before an invoice gets paid: the purchase order (what we ordered), the receiving document (what arrived), and the invoice (what the supplier is charging). Industry data places unrecovered losses at 1% to 2% of total accounts payable spend for organizations without systematic matching. On a $10M AP run that is $100K to $200K a year in quiet leakage: duplicate payments, price creep, quantity shorts, and fraudulent invoices against vendors that do not exist.
The classic small-shop failure is the "two-way match": PO matches invoice, nobody confirms the goods arrived. Parts get booked paid that never showed up. Or worse, the receiving document gets updated by the same person who approves the invoice.
TransAlta, a Canadian power company, lost $24M on a single copy-paste misalignment in a procurement spreadsheet. Ten percent of that year's profit, gone to a row error. Small manufacturers do not lose that number in a single event, but the structural leak is the same. Spreadsheet-driven procurement carries an error tax nobody is tracking.
10. Keeping the whole procurement file in one person's head
The last mistake is the one that kills enterprise value. At most small manufacturers, procurement knowledge lives in one person. That person knows which supplier will make a small run without a setup charge, who to call when a bar-stock shipment is short, which freight broker will hold a truck an extra day, and what price floor each supplier will actually accept. None of it is written down.
If that person gets hit by a bus, so does the shop. If that person retires, the replacement buyer spends two years re-learning relationships and paying rookie prices in the meantime. If the owner goes to sell the business, the buyer reads the arrangement as exactly what it is: a single point of failure in a critical function, discounted accordingly.
Tribal knowledge in procurement is one of the most common reasons a business gets a lower multiple in diligence than the owner expected. An EV/EBITDA multiple is paid for operational resilience, and procurement dependence on one person is the opposite of that. The fix is documented supplier files: approved-vendor list with qualifications, price history, terms, lead times, quality history, and contact of record. It is a weekend of work for a buyer with ten years of memory. It is worth six figures at exit.
A rough maturity ladder
Most US small manufacturers sit on the bottom two rungs. Climbing one rung pays back in gross margin within two quarters.
| Level | Behavior | What it looks like |
|---|---|---|
| 1. Reactive | POs driven by stockouts and shop-floor requests. No approved vendor list. | Spreadsheet and email. Prices drift. No one can produce a supplier scorecard. |
| 2. Transactional | Formal PO process. Three-way match. Approved vendor list. | The procurement function is the PO workflow. Nobody is negotiating. |
| 3. Tactical | RFQs on new parts. Supplier scorecards on the top 20 suppliers. Negotiated agreements on the top 80% of spend. | Savings are measurable. Lead-time variance is being managed. |
| 4. Strategic | Should-cost models, dual-source strategy by category, supplier development, TCO-driven awards. | Procurement is a board-level conversation. Supplier relationships are multi-year. |
Most of the climb from Level 1 to Level 3 is process and discipline, not software. Inventory visibility, accurate job costing, and clean BOM data make it possible to run Level 3 without an enterprise-grade procurement suite. A shop that knows its true costs and its true inventory can make procurement decisions an outsider would recognize as strategic.
Why procurement maturity shows up in the selling price
Buyers of manufacturing businesses care about procurement for the same reason they care about job costing and on-time delivery: it is a leading indicator of how the business will perform after a transition. During diligence, a quality-of-earnings team will ask for trailing-twelve-month spend by supplier, price history on the top 20 parts, supplier concentration, and any long-term agreements in force. A shop that can produce those numbers in a day looks fundamentally different from a shop that has to assemble them from email.
Supplier concentration is evaluated the same way customer concentration is. A top supplier above 25% of COGS is a yellow flag. A sole source without a qualified alternate is a red flag. Both show up in the price and the structure of the deal. A procurement function that has documented dual-source qualifications, supplier scorecards trending in the right direction, and contracted pricing on the majority of spend reads as a business where risk has already been priced. That is worth a turn of EBITDA at the size most US manufacturers sell at. Owners who want to understand the math should read what determines enterprise value for manufacturers.
Frequently asked questions
What is procurement, in plain terms?
Procurement is the strategic process a business uses to buy the goods and services it needs from outside suppliers. For a manufacturer, that means raw materials, components, consumables, MRO supplies, tooling, subcontract work, and services. Procurement is responsible for the decision of who supplies what, at what price, under what terms, and with what accountability when they miss.
What is the difference between procurement and purchasing?
Purchasing is transactional: issue a PO, receive the goods, pay the invoice. Procurement is strategic: choose the suppliers, negotiate the contracts, manage the performance, and own the total cost. A purchasing clerk executes decisions. A procurement function makes them. Most small manufacturers have purchasing but call it procurement, and the savings that procurement would deliver go unclaimed.
What are the stages of the procurement process?
Six stages, known collectively as source-to-pay. Demand identification, sourcing, contracting, purchase order issuance, receipt and invoice matching, and supplier performance management. Each stage has its own artifacts (specifications, RFQs, contracts, POs, receiving documents, invoices, scorecards). Each is usually owned by a different person in a small shop, which is why things fall between them.
How much do procurement mistakes actually cost a small manufacturer?
A useful anchor: maverick spend typically runs 20% to 30% of indirect spend leakage, tail spend is 10% or more of total spend with only 4% of companies managing it, AP errors without a three-way match leak 1% to 2% of accounts payable, and over-specified parts cost 20% to 100% more than industry-standard equivalents. Stack those on a shop doing $20M in revenue and the exposure is routinely six figures a year in recoverable margin. The savings are not in finding a cheaper supplier. They are in fixing the process that leaks to every supplier.
Does a small shop need procurement software?
Usually no, at least not at the start. Most of the climb from reactive purchasing to a tactical procurement function is a written process, a short approved-vendor list, and a scorecard somebody updates monthly. ERP systems that already exist in most manufacturers (especially purpose-built manufacturing ERP) cover the three-way match, PO approval, and spend analysis for free. Dedicated procurement suites start paying back around the point where indirect spend is large enough to staff a full-time buyer and broad enough to benefit from catalog and contract automation.
Who should own procurement in a 50-person manufacturer?
At that size, there is usually no full-time procurement manager. The function needs to be owned by someone senior, typically the controller or operations manager, with buyer-level execution below them. What matters is that one person is accountable for supplier performance, price trends, and savings. If three people split the responsibility, the function does not exist.
Conclusion
Procurement at a small US manufacturer is usually the gap between what the business is earning and what it could earn. The ten mistakes above are not exotic. They are the default state when nobody owns the function, and they cost more now than they did five years ago because tariffs, tight supplier capacity, and post-COVID disruption have raised the price of every weakness.
Most of the fixes are cheap. A written PO process. A one-page scorecard per top supplier. A should-cost model on the top 20 parts. A qualified alternate on every sole source. A documented buyer file. None of this requires a procurement officer or a seven-figure software purchase. It requires a decision that procurement is a function, not a clerking task.
Want to see what clean procurement data looks like when it is tied to inventory, job costs, and supplier performance? Book a demo and we will show you how WorkCell gives small manufacturers the procurement discipline of a much larger shop.