Your shipment volume grew over the last 2 years. Your cost to serve per shipment grew right along with it.

The volume, you earned. Your customers have spent the last 2 years redrawing their sourcing maps. In McKinsey’s 2025 survey of global supply chain leaders, 43 percent said they plan to shift more of their supply chain footprint to the United States within 3 years, a 25 percentage point jump in a single year, and 38 percent plan to reduce their presence in China. Production that had lived at one address for a decade suddenly had a new one. You competed hard for the new lanes and won your share. For a while it felt like the market was finally paying you back.

Then the quieter number moved the wrong way. Not freight cost. Rate movement you can explain in your sleep. Cost to serve is what your own operation spends to handle each shipment: the documentation, the exceptions, the milestone updates, the billing, the collections. Volume rose, and the cost of serving each unit of it rose with it. The win did not land.

You responded the way the industry taught you to respond to a shoring event. You opened a desk closer to the new origin. You added heads on the lane. Maybe you stood up a nearshore team to absorb the documentation load, because that is what everyone was doing and the labor math looked clean.

Every one of those moves was reasonable. None of them was wrong on its own terms. The lane is covered. The customers are served. And the cost is still there, sitting in the per-shipment number, quarter after quarter.

So here is the question worth sitting with: why did a playbook that works so well for your customers fail inside your own operation?

When your customer relocates production, an asset moves. A factory. A supplier. A distribution node. Assets have addresses. You can pin them on a map, model the duty impact, reroute the freight, and price the lane. The entire vocabulary of shoring, offshore, nearshore, onshore, reshore, was built for exactly that problem: where should the thing be made?

When your operation absorbs that change, nothing with an address moves. What moves is work. Entries to file, documents to cut, exceptions to chase, milestones to update, invoices to issue and collect. Work is not an asset. It is a chain of handoffs, and handoffs do not have addresses. They have owners. Or they do not, and that is where the cost hides.

Here is the assertion, and you are free to argue with it: your cost to serve rose because you answered a process question with a footprint answer. You solved for where at the exact moment your operation was asking who owns this handoff, and how is it governed. The desk near the origin covered the geography. Nobody redesigned the ownership. So every exception still travels the same unowned path it always did, only now there are more of them. Unowned exceptions come back as rework, and rework is cost to serve.

The market’s own numbers show the gap. When ISG surveyed 368 executives responsible for business process outsourcing in 2024, 68 percent named cost reduction as their top motivator. The average realized saving over in-house operations came in at 15 percent, and only 38 percent rated the savings they achieved as very good or excellent. Set that 15 percent against the per-seat savings your business case promised, and the distance between the two is the story. ISG’s own analysts suggest why: providers are typically measured on process outcomes rather than broader business results. In plain language, everyone is watching the tasks. Nobody is accountable for the handoffs.

You could blame rates. You could blame mix, or the fact that new customers are always messier than the ones you have had for years. Sometimes that is even true. But rates normalize and mix settles, and in my experience the operations that carry this problem for years all share the same signature: the volume was won, the footprint responded, and the ownership layer was never touched.

There is a name for that ownership layer, and it is probably not the one you think, because the market has spent years teaching the wrong definition. Search the term and you will find bestshoring defined as picking the best location to move manufacturing, production, or IT. Goods vocabulary again. The suffix survived. The subject got lost.

Bestshoring was never a goods discipline. It is the design layer for the work itself: where your work should live, who should do it, and how it should be organized. Your customers’ sourcing decisions are demand-side events, and you serve them well. This layer governs the supply side you own, the operational backbone of documentation, customer service, exceptions, and settlement that every shipment runs through. The Bestshoring Architecture™ puts those decisions in a deliberate order, and geography comes last in the stack, not first.

This is not theoretical. When we rebuilt a nearshored shipment processing operation around a one file, one owner model, the location moved too, to Bogota, but location was the last decision made, not the first. Ownership was designed before a single seat moved. The results, published in full in the case study: customer response times improved 20 percent, KPI performance rose 70 percent, labor savings exceeded 60 percent, and the operation delivered 1.2 million dollars in annual cost reductions with breakeven inside the first year.

Read that sequence again. Ownership first. Governance second. Geography last. The savings held because the handoffs had owners before they had a new address.

That is what your win is supposed to look like. Ownership assigned, handoffs governed, and the volume you earned from your customers’ footprint moves arriving as margin instead of leaking out as rework. Volume up. Cost to serve per shipment down. The win, finally landing where you can bank it.

1. For the 5 largest customers you gained in the last 2 years, can you name the single owner of exception handling on each lane? Not the team. The owner.
2. When cost to serve rises on a lane, is your first diagnostic a map or an ownership chart?
3. If a major customer moved production again next quarter, would your operation absorb it by design, or by heroics?