Tariffs Are a Margin Problem, Not a Sourcing Problem

The Question Everyone Is Asking, and the One They Should Be

Walk into any small or mid-market manufacturer this summer and there is a version of the same conversation happening in a conference room somewhere: where do we move production, and how fast can we do it?

It is the wrong question. Not because sourcing does not matter, but because for the overwhelming majority of manufacturers, tariffs are not destroying demand. They are quietly eating margin, one purchase order at a time, in a way that never shows up as a crisis and therefore never gets a project team assigned to it.

The damage is arriving as margin compression rather than demand destruction, and that distinction changes everything about the correct response.

The damage is arriving as margin compression rather than demand destruction, and that distinction changes everything about the correct response. A demand problem is solved with sales and capacity. A margin problem is solved with visibility, speed, and the ability to model a decision before you commit to it. Most manufacturers are staffing for the first problem while losing money to the second.

What the Data Actually Says About Reshoring

Start with the assumption everyone is operating on, because it does not survive contact with the evidence.

Trade uncertainty is the defining concern of the moment. Deloitte's 2026 manufacturing outlook found that 78 percent of manufacturers cited it as their top concern, and that input costs are expected to rise by an average of 5.4 percent over the coming year. Survey reporting through 2026 has more than half of small and mid-size businesses citing a greater tariff impact than they felt twelve months ago. The pressure is real and it is intensifying.

But the popular conclusion drawn from that pressure is wrong. According to Institute for Supply Management data, 64 percent of manufacturers do not intend to bring production to the United States to avoid tariff costs. That is not hesitation or denial. It is arithmetic. Domestic capacity is expensive, skilled labor is scarce, and building or retrofitting a plant takes twelve to twenty-four months before it produces a single part. A tariff schedule that may change before your concrete cures is a poor foundation for a capital project.

What manufacturers are actually doing is more pragmatic and less headline-friendly: qualifying alternative suppliers in lower-tariff regions, front-loading inventory where they have the storage and the working capital to carry it, and renegotiating terms. Reshoring, where it happens, is a multi-year strategic bet that happens to be helped by tariffs. It is not a tariff response.

Margin Compression Is the Real Damage

Here is what the slow version of this problem looks like from the inside.

A part you have bought from the same supplier for nine years now lands at a different cost, and the tariff line is only part of it. Freight changed. The broker's classification changed. Your customs treatment on a subassembly turned out to be different from the treatment on the finished good. None of these are large enough individually to trigger an alarm. Together they move a product family from a healthy margin to a marginal one, and you find out at quarter close.

Meanwhile the quoting process runs on last year's landed cost, because that is the number in the system. Every quote you win at the old number is a quote you would have priced differently if you had known. The revenue looks fine. The margin does not, and by the time the variance surfaces you have already committed to twelve months of deliveries.

There is a second, subtler cost that rarely gets counted. Tariff mitigation consumes planning bandwidth that would otherwise go toward capital investment in efficiency, infrastructure, and growth. The senior people who should be improving throughput are instead rebuilding cost models in spreadsheets. The tariff bill is visible. The opportunity cost of the attention it absorbs is not.

Your ERP Already Holds the Answer, and You Probably Cannot Get to It

The frustrating part is that almost every fact needed to manage this well is already sitting in your ERP. Every purchase order, every supplier, every part cost, every lead time, every customer price, every margin by product family. The data is there.

What is usually missing is the ability to ask it a question that spans more than one table and get an answer the same day. Landed cost lives in one place, tariff classification in another, supplier lead times in a third, and the quote that depends on all three lives in a spreadsheet on someone's desktop. So the analysis gets done manually, once a quarter, by whoever has the time, using an export that was already stale when it was pulled.

That gap between having the data and being able to act on it is the whole problem. It is also, in 2026, an unusually solvable one. You do not need a new ERP to close it. You need a layer that reads the system you already own, models the scenarios you actually face, and pushes the answer to the people making pricing and sourcing decisions before they make them rather than after.

The Four Moves That Actually Work

The manufacturers handling this well are not doing anything exotic. They are doing four unglamorous things consistently.

The first is landed cost at the line level, kept current. Not an annual standard cost that everyone knows is wrong, but a duty-inclusive, freight-inclusive, classification-aware cost that updates when the inputs change and flows into quoting automatically. This single change surfaces more margin leakage than any other, because it converts a quarterly surprise into a daily fact.

The second is scenario modeling before commitment. If this supplier moves to that country, what happens to landed cost, lead time, working capital, and margin by product family? Deloitte notes that a majority of trade professionals are now using technology for exactly this: trade route analysis, risk identification, cost savings, and scenario modeling. The manufacturers who can answer that question in an afternoon make better decisions than the ones who need three weeks, and they make them while the option is still open.

The third is supplier qualification as a standing capability rather than an emergency project. This is sourcing work, which may look like a contradiction in an article arguing that sourcing is not the problem. It is not. The point is not to switch suppliers; it is to have a qualified alternative already through your quality process, so that switching becomes a decision you can make in weeks rather than a program you start from zero under pressure. Optionality is a margin instrument, not a sourcing strategy.

The fourth is disciplined inventory positioning. Front-loading inventory is a legitimate hedge, but it consumes working capital, and for capital-constrained small and mid-market companies that financing has a cost that has to be modeled against the tariff exposure it offsets. Buying ahead without that math is trading a known cost for an unknown one.

The Capability That Outlasts the Tariff

Notice what all four of those moves have in common. None of them depends on any particular tariff schedule staying in place.

That is the argument for treating this as an operating capability rather than a one-time response. Tariff regimes change with administrations, negotiations, and court rulings. The ability to recalculate landed cost overnight, model a sourcing change in an afternoon, and reprice with confidence is valuable in every one of those futures. It is equally valuable when the disruption is a port closure, a supplier bankruptcy, a raw material shortage, or a currency move.

There is likely a commercial argument too. Deal-side reporting in 2026 describes strategic buyers and private equity groups weighting operational resiliency, customer visibility, and supply chain control more heavily than they used to when they evaluate manufacturers. We would not tell you to build a resiliency program for the valuation. But if a sale or a recapitalization is anywhere in your horizon, the work you do to protect margin is unlikely to hurt you in diligence.

The manufacturers who come out of this period stronger will not be the ones who guessed the tariff schedule correctly. They will be the ones who built the ability to answer sourcing and pricing questions faster than their competitors, and kept it.

The Bottom Line

Tariffs in 2026 are not primarily a sourcing crisis for small and mid-market manufacturers. They are a margin problem that arrives quietly, compounds across thousands of transactions, and is usually discovered at quarter close rather than at the moment of decision. Answering it with a plant relocation is expensive, slow, and for most companies the wrong instrument entirely.

The higher-return response is to make landed cost visible and current, make scenario modeling fast enough to inform decisions rather than explain them, keep alternative suppliers qualified in advance, and price with real numbers. All of that is built on data you already own, in a system you have already paid for.

At Cherry Street, this is the work we do. We help manufacturers get at the cost, supplier, and margin data trapped in their ERP, build the scenario models that turn a sourcing question into a same-day answer, and put those numbers in front of the people quoting and buying. If your tariff response so far has been a spreadsheet and a lot of senior attention, there is a better version of it that does not require replacing anything.