What a Sourcing Strategy Actually Decides
This article is written for U.S. sourcing, procurement and supply-chain managers who buy manufactured parts, components and finished goods from outside the United States. International manufacturers can read it backwards, as the logic a buyer applies when deciding how many suppliers to hold, where to place them and what a quote really costs.
A sourcing strategy is not a list of approved vendors. It is a small set of decisions that, once made, determine most of your cost, risk and flexibility for years: how many suppliers you hold per part, where they sit, how long your supply lines are, which cost basis you compare them on, and what triggers a change. Everything else is execution. Teams that skip the decisions end up with a supply base assembled part by part, usually by whoever had the most urgent shortage that quarter.
Single, Dual and Multi-Sourcing
Supplier count is the first structural choice, and each option buys something real at a real price.
Single sourcing puts all volume with one manufacturer. It gives you the strongest price position, amortizes tooling once, and builds an engineering relationship deep enough for the supplier to contribute to design and cost reduction. It also means a fire, a labor dispute, a capacity reallocation or a quality escape at one plant becomes your problem immediately, and it weakens your leverage once switching would take months.
Dual sourcing splits volume between a primary supplier and a secondary one carrying a smaller share. You pay twice, in duplicate tooling and qualification cost and in a weaker price on split volume. What you get is a live alternative, and the distinction that matters is between a second source and a second quotation. A supplier you qualified two years ago and never bought from again is not a second source. Keeping one warm means recurring volume, periodic requotes, current samples and a tooling condition you have verified.
Multi-sourcing spreads a category across several suppliers. It works where the item is standard and switching cost is close to zero, and it is slow and expensive for anything engineered to your drawing.
A workable default: single-source where tooling is heavy and the part is proprietary, dual-source anything whose absence would stop your production or breach a customer commitment, and multi-source where the market already gives you interchangeable options.
Regional Diversification, Distance and Concentration
Concentration risk is rarely visible at the supplier level. Two approved suppliers in the same industrial cluster can share a port, a labor market and, more often than buyers realize, the same sub-tier source of raw material. Two suppliers with one common bottleneck is one supplier with extra paperwork, so for parts you cannot afford to lose, map at least one tier below your direct supplier.
Regional diversification means holding capability in more than one manufacturing region so that a disruption, a capacity squeeze or a change in trade conditions in one place does not remove your whole supply. Be precise about the limits of that argument. Tariff exposure, duty rates, country-of-origin determination and eligibility under trade programs depend on product classification, the manufacturing steps actually performed and your commercial structure, and they are questions for qualified customs brokers and trade counsel rather than assumptions to build a strategy on.
Distance is the other half of the decision. Shorter supply lines reduce transit time, in-transit and safety stock and the cash locked in the pipeline, and they make engineering changes and problem-solving faster. Longer lines often reach deeper specialized capability and a better unit cost position. The honest comparison is not unit price against freight; it is total cash in the pipeline and speed of response against unit cost. Products with volatile demand or frequent design change belong on shorter lines, while stable, forecastable, low-mix products tolerate long ones well, and many mature programs run both. How much inventory each structure obliges you to hold, and where, is as much a 4PL and warehousing question as a purchasing one.
Total Landed Cost, Not Unit Price
Unit price is the least reliable number in a sourcing comparison. A defensible landed cost model includes, at minimum:
- Ex-works or FOB price, plus tooling amortized over a realistic volume.
- Inland freight to the origin port, international freight, cargo insurance, port charges, drayage and domestic freight.
- Import duties, fees and brokerage, treated as product-specific figures confirmed with qualified customs professionals rather than a general percentage.
- Warehousing, handling, and the carrying cost of in-transit inventory and of the safety stock that lead time and its variability force you to hold.
- Payment terms and their working-capital effect, plus expediting you expect to use.
- Cost of quality, meaning inspection, sorting, rework, scrap and warranty exposure, and the management cost of travel, engineering support and internal time.
The last two are almost always missing when a proposal reaches a steering committee, and both are usually larger than the unit-price saving that justified the decision. Treat landed cost as a range, state your assumptions, and revisit it when freight, currency or trade conditions move.
Matching the Strategy to the Part
One policy across all spend produces both over-managed commodities and single-sourced critical parts nobody noticed. Segment instead, using value at risk, engineering complexity, switching cost, demand volatility and the number of alternatives. Four profiles cover most industrial spend:
- Engineered, tooling-heavy, few alternatives. Go deep with one supplier, secure continuity and tooling ownership contractually, and develop a second source deliberately, not under pressure.
- Critical but technically standard. Dual-source across different regions and keep real volume flowing to both.
- Competitive commodity. Multi-source, requote on a defined cycle, and spend your effort on specification rather than relationships.
- Low value, high transaction count. Consolidate to reduce administrative load; redundancy belongs at the category level.
Then write your triggers down in advance: a sustained delivery miss, an adverse quality trend, a refusal to support a capacity increase, a change of ownership, or a material shift in freight or trade conditions. Strategies that change only during a crisis are not strategies.
Building the Supply Base You Actually Need
Strategy is the easier half. The harder half is finding manufacturers who can genuinely meet the specification, verifying it before you depend on them, and running the commercial process well enough to compare them fairly. That work is supplier identification, qualification and RFQ management, and it is what Expanvia's Sourcing capability does in both directions: connecting U.S. demand with global manufacturing, and helping capable manufacturers reach U.S. buyers. Where customs, trade compliance, legal or tax questions arise, we coordinate with qualified professionals rather than answering them.
If you are rethinking a category and want to pressure-test the structure against your own parts, volumes and risk tolerance, start with Let's Talk.