North American Automotive Wire Harness and Cable: Supply Base and Application Map

The wire harness is the part of the vehicle that everybody depends on and nobody sees. It is also the part that has resisted automation longer than any other, and that single fact explains almost everything about how this market is organized.

A mid-size vehicle harness is two to three miles of wire, three to five thousand terminals, several hundred connectors, and somewhere between twenty and forty hours of human hands laying wire onto a board, seating terminals, taping bundles, and testing continuity. Body shops are robotic. Paint shops are robotic. Harness assembly is people at benches, and it has stayed that way because the product is a branched three-dimensional assembly whose geometry changes with every option code on every build order. An automated cell can cut, strip, and crimp a lead. It cannot lay a forty-branch harness onto a board and route it around a bulkhead.

So the work went where labor is. In North America that means Mexico, and it has meant Mexico since the late 1980s, well before USMCA turned the location into a tariff question as well as a wage one. The assemblers maintain engineering and program management offices in metro Detroit and in Kentucky, and they build in Chihuahua, Coahuila, Durango, Sonora, Nuevo Leon, and Ciudad Juarez. A US address on a harness company is a sales office. The plant is south of the border.

Four companies hold most of the market globally and in North America: Aptiv, Yazaki, Sumitomo Electric Wiring Systems, and the Lear E-Systems group. Below them sits a second rank of European and Japanese assemblers running Mexican plants against specific programs, several of which have changed ownership in the last three years in ways that are still working themselves out. The concentration is real, it is stable, and it is not the interesting part.

The interesting part is that this category is splitting in two. Low-voltage harness is a mature business under continuous cost pressure with a structural threat on the horizon: zonal architecture, which reduces total harness length by replacing long point-to-point runs with short runs to local controllers. High-voltage harness is a different product with a different bill of materials, different standards, different testing, and margins that low-voltage assembly has not seen in twenty years. It did not exist in volume ten years ago. A supplier entering this market needs to know which of the two businesses it is entering before it books a single meeting, because they share a name and almost nothing else.

Copper is the other thing to understand up front. It is the single largest material cost in a harness, it is bought against an exchange index and passed through to the OEM on a formula, and that pass-through means a copper price move is a working capital problem for the assembler rather than a margin problem. What it also means is that every conversation about reducing harness cost eventually becomes a conversation about using less copper, which is why aluminum substitution has been attempted for twenty years and why zonal architecture is being pursued now. Both are covered in Section 10.

For a materials supplier, the practical read is this: harness assembly itself is a labor business with thin margins and four entrenched incumbents, and it is not a realistic entry point. The components are. Tape, conduit, clips, grommets, insulation compounds, and the resins under connector and fuse box housings are real, addressable, and in several cases genuinely contestable. Section 9 is about exactly that.

This lens uses the industry tier labels deliberately, and the definitions differ from the other automotive lenses because the decision rights are distributed differently.

The OEM specifies the connector system and then steps back. This is the structural fact that governs everything else. During vehicle development the OEM electrical architecture group approves a connector system: a family of housings, terminals, seals, and mating parts from a named supplier, validated to the OEM's own test specification. That approval is made years before production and it is held at the OEM, not at the assembler. When the harness goes out to bid, the drawing already names the connectors. An assembler cannot substitute them, and a connector maker that is not on the approved list cannot sell into the program at any price. The window to win a connector position is the architecture phase, and it closes quietly.

Tier 1 is the harness assembler, and it is a labor business. The assembler buys wire, terminals, connectors, tape, conduit, and clips, and sells a finished, tested, sequenced harness to the vehicle plant. Its value add is engineering the routing, managing a bill of materials with thousands of lines and hundreds of option-driven variants, and converting labor into a product that arrives in build sequence. Margins are thin by automotive standards, and the competitive variables are wage rate, plant efficiency, engineering responsiveness, and program management. Four companies hold most of it: Aptiv, Yazaki, Sumitomo Electric Wiring Systems, and Lear E-Systems.

Three of the four Tier 1s are also Tier 2s, and that is not a footnote. Aptiv, Yazaki, and Sumitomo all make their own connectors and terminals, and all three sell those components to competing assemblers. That means the company quoting against you on a harness program may also be the company you have to buy your connectors from, and the OEM approved list frequently names one of them. Lear is the least vertically integrated of the four and buys more of its component content, which makes it the most open door for a component supplier and is worth knowing.

Tier 2 is the component maker, and it sells into the assembler against the OEM's drawing. Connector and terminal makers, TE Connectivity, Molex, Amphenol, JAE, and the captive component arms of the assemblers, live or die on approved-list position. Tape, conduit, and clip converters live differently: those parts are frequently specified by performance class rather than by part number, which leaves the assembler with genuine discretion over the supplier. That discretion is the single most important commercial fact in this lens for a new materials supplier, and it is why tape and conduit are the recommended entry point in Section 9.

Tier 3 is copper and insulation resin. Copper rod and drawn wire producers, copper strip suppliers to the terminal stampers, and the PVC, polyethylene, polypropylene, and engineering resin producers under insulation, conduit, clips, and connector housings. The copper conversation is not a price conversation, because the metal is indexed and passed through. It is about conversion cost, capacity assurance, and increasingly about recycled content, which is a genuine differentiator in a category where secondary copper is abundant and well characterized.

Distribution here is not how production works. Heilind sells interconnect in sample and prototype quantities. Waytek sells harness supplies to upfitters and low-volume builders. Anixter and Wesco sell bulk cable at national scale, mostly into facilities and infrastructure rather than into vehicles. All three are real businesses and none of them is how a production car harness gets built. A supplier that gets excited about a distributor relationship in this lens has usually misunderstood which market it is in.

The flow of money and the flow of decisions run in different directions. The OEM pays the assembler. The assembler pays the component maker. But the decision about which component maker gets paid was made at the OEM, sometimes five years earlier, by an engineer who never saw a purchase order. Selling into this category means understanding that the purchase order and the decision come from different buildings, and calling on only one of them is calling on half the market.

Every light vehicle OEM building in North America appears in this lens, because every vehicle has a harness. What differentiates them here is narrower than the company profile: which connector systems they have approved, how far they have gone toward zonal architecture, and whether their high-voltage content is engineered in house or bought as a system.

The Detroit three specify heavily and buy locally engineered. GM, Ford, and Stellantis each maintain electrical architecture groups that approve connector systems, write wire specifications, and hold the routing drawings. Aptiv's position in North America is rooted here through its Delphi heritage: the Delphi connector systems are deeply embedded in GM architecture in particular, and that embedding is decades old and very hard to displace. All three are also the OEMs pushing hardest on harness cost, because their volumes make a dollar per vehicle a real number.

The Japanese transplants bring their own component ecosystem with them. Toyota, Honda, Nissan, Subaru, and Mazda build in North America against architectures approved in Japan, which means Sumitomo, Yazaki, Furukawa, and JAE connector systems arrive on the drawing already approved. This is the clearest example in the atlas of a supply base that travelled with its customer. A North American component supplier trying to enter a transplant program is not competing on price against an incumbent, it is trying to get onto a list that was closed before the plant was built. PBT usage runs measurably higher on these programs than on Detroit-domestic ones for the same reason: the material preference travelled too.

The German transplants bring European specifications, including the halogen-free ones. BMW in Spartanburg, Mercedes in Alabama, and Volkswagen in Chattanooga build to architectures written in Germany, and those specifications carry material rules that North American regulation does not require. Halogen-free insulation is the most consequential: it is not a US regulatory requirement and it is arriving in the US supply base anyway, program by program, because the specification was written at a design centre in Munich or Wolfsburg. A supplier watching for where halogen-free insulation becomes volume should watch these three.

The EV natives are where the high-voltage content is densest and the architecture is least settled. Tesla has been the most aggressive in the industry about reducing harness length and content, and its architectural choices have been studied and partially copied by every OEM in this list. Rivian and Lucid engineer more of their high-voltage systems in house than a legacy OEM does, which shortens the path from a materials supplier to an engineer who can actually decide something. Scout, still pre-production, is the rarest opportunity in this lens: an architecture being written now, where a connector system and a wire specification are still open questions.

Everybody is somewhere on the zonal path and nobody is finished. Zonal architecture replaces long point-to-point runs from a central body controller with short runs to local zone controllers connected by a high-speed backbone. It shortens the harness, reduces copper mass, and simplifies the variant explosion that makes harness bills of materials so expensive to manage. It also requires a software and controller architecture most OEMs are still building. Every OEM on this list has announced a version of it, the timelines have all slipped at least once, and the transition will be gradual and platform by platform rather than sudden.

The full roster of eighteen OEM positions, the connector approval politics behind each one, and where each sits on the zonal transition, is in the gated portion of this report.