SKF Precision Automotive Parts: In-House Tooling vs. Outsourced Dies – A Procurement Manager’s Reality Check

When It Comes to Automotive Parts, “Make or Buy” Isn't the Only Question

I've been handling procurement and engineering support for automotive tier suppliers for about 12 years now. In that time, I've personally signed off (and sometimes regretted) decisions that flushed roughly $80k in wasted budget down the drain. I now maintain our internal checklist for die and mold sourcing, which has caught 47 potential errors over the last 18 months.

One of the most persistent debates I run into is this: Should you invest in your own tooling and die-making capabilities (like SKF does for certain precision lines), or is it smarter to outsource to specialized die shops?

This isn't a simple “A vs. B” story. Both paths work. Both can fail spectacularly. But the industry has evolved a lot since I started in 2017. What was best practice in 2020 may not apply in 2025. Let me walk you through the comparison based on specific dimensions, not generic opinions.

---

The Comparison Framework: Precision vs. Volume Flexibility

Before we jump into the details, let me clarify the baseline. We're comparing two scenarios:

• A) In-House Tooling (Vertical Integration): Think SKF's own die and mold shops. The company controls the entire design, fabrication, and maintenance of tools. This is common for core product families like CV joints (e.g., SKF 641091) and wheel bearing assemblies.
• B) Outsourced Die & Mold Sourcing (Specialist Suppliers): You send your part design to an external tool maker. They build the progressive die or casting mold off-site. You then use that tool at your own stamping/forging line.

I'll split the comparison into three dimensions where I've seen the most painful (and avoidable) failures: Tooling Lead Time & Lifecycle Cost, Tolerance Repeatability, and Design Change Agility.

---

Dimension 1: Tooling Lead Time & Lifecycle Cost (The Hidden Budget Trap)

The In-House Argument (and the Catch)

Proponents of in-house tooling often cite control over lead times. “We can skip the queue, prioritize our own projects.” That's true—to a point. I've seen SKF's internal die shop turn around a repair for a progressive stamping die for a critical chassis bracket in 3 days when an external shop quoted 2 weeks. That kind of speed saves production downtime (which can cost $500 an hour or more).

However, I learned the hard way that “in-house” doesn't automatically mean “cheap.” In Q4 2023, I had to authorize a budget for a new in-house tooling cell for a multi-process part (stamping + CNC + some extrusion). The initial capex was eye-watering. The justification was that we'd save 20% per die over the next 3 years. I bought in.

Then the real costs hit: machine utilization rate was only 60% for the first 9 months. The CNC programming for a complex aluminum extrusion die required a specialist we had to hire externally anyway (which, honestly, I should have budgeted for). Total cost? About $23k over the original quote. The external shop would have been 10% more expensive per die, but with zero capex overhead. Lesson learned: total cost of ownership isn't just the die price—it's the utilization rate and the hidden cost of idle capacity.

The Outsourced Reality (and the Counterpoint)

Outsourcing often looks cleaner on the P&L. You pay for a completed tool. No asset depreciation. But here's the catch I discovered on a $3,200 order of forged connecting rods: the external die maker was 2 weeks late, and their design didn't account for our specific press tonnage (an overconfidence fail on my part—I skipped verifying the shut height). The die arrived, we installed it… and it didn't fit the press. $890 rework + a 1-week production delay.

Bottom line on Dimension 1: In-house wins if you have >70% utilization on your tooling capacity AND can absorb the capex. Outsourced wins if you need variable capacity without capital commitment. (I can only speak to mid-size B2B operations like ours—if you're a low-volume prototype shop, the calculus might be different.)

---

Dimension 2: Tolerance Repeatability & Quality (The “Perfect” Part Myth)

The In-House Edge (and Its Limits)

In-house tooling gives you direct control over grinding, heat treatment, and die spotting. For a component like the SKF 641091 CV joint (which needs sub-micron tolerances on the raceway surface), that control is non-negotiable. The in-house team can iterate on the die polish in real-time, without sending samples back and forth.

But here's a truth that surprised me early in my career: in-house tooling doesn't automatically mean better parts. In September 2022, our own die shop made a progressive die for a stamped bracket. The first 200 parts looked great. Then the die started galling (aluminum pick-up on the steel) because our coolant flow wasn't optimized for the new alloy. We'd skipped a final simulation step (a process gap). Three days lost, and we had to scrap 50 parts. Cost? About $1,200.

Industry standard tolerance for high-speed stamping is often 0.020mm for critical dimensions. Can an external shop hit that? Absolutely—if you pay for precision and if they specialize in that alloy.

Outsourced Reliability (When It Works)

For a simple, high-volume part—think a non-critical heat shield bracket—outsourced dies are often more reliable than in-house. Why? Because the external shop lives and dies by their die performance. They've made 500 similar dies. The in-house team might make 12 per year. Experience matters. I've had an external shop deliver a die for an Eaton differential component that ran for 80,000 stampings without any unscheduled maintenance. That's a win for total cost of a part.

Bottom line on Dimension 2: For precision-critical parts (wheel bearings, CV joints, high-stress forged components), in-house control is hard to beat. For standard parts with moderate tolerances, a specialist die shop often delivers better initial quality and longevity (based on recent data—this might change as AI-driven die design evolves).

---

Dimension 3: Design Change Agility (The “Oops, We Need a Revision” Moment)

This is where I've seen the most dramatic divergence between the two approaches. A customer (let's just say it was a tier 1 supplier for a German OEM) came back with a change request on a stamped part. The flange angle needed to shift by 8 degrees. Simple change on the CAD model. But how fast can you modify the die?

In-House: Fast, But at a Cost

In-house tooling can handle that change in days—maybe a week, depending on workload. The die is physically in your shop. You can do a weld-up and re-cut the die surface. That's a massive advantage for prototypes or low-volume production runs (e.g., 50 pieces for a pre-production build).

But here's the twist: I once approved an urgent engineering change on a complex progressive die (for a part that ultimately went into a an e46 muffler bracket replacement). The in-house team welded up the die in 2 days. We approved the change. Then we realized the new design didn't play well with the existing draw stage. The die cracked after 150 parts (unfortunately). Total cost of the revision was $3,400 for the re-cut plus scrap.

Outsourced: Slower, but with a Built-in Gate

Outsourced dies mean you have to ship the tool back or build a new one. That's slower (usually 2-4 weeks). But here's the counter-intuitive benefit I've observed: the external shop automatically reviews the change for manufacturability. They might say, “Hey, if you move that radius by 0.5mm, we can use an existing insert and save you $1,000.” In-house teams sometimes miss that because they're too close to the part.

I once had an external die maker in Italy who caught a design flaw on an aluminum extrusion for a heat sink component. That catch saved us from a major failure down the line. They earned my respect (and future orders).

Bottom line on Dimension 3: In-house wins for speed on minor tweaks (with risk of local blindness). Outsourced wins for thorough design review (with time penalty).

---

So… Which Path Should You Take?

Here's the uncomfortable truth: It depends on your part's criticality and your internal team's bandwidth.

Go in-house (or invest like SKF does) if:

• Your parts have tight tolerances manufacturing tolerance range of Rockwell 48 for the die steel (as of 2025, a standard for high-wear applications) .
• You need rapid iteration for R&D or prototypes.
• You have the volume to keep your tooling capacity utilization above 70%. If you're unsure, benchmark against your internal machine hours—I use a simple spreadsheet that tracks actual vs. available hours. It's eye-opening.

Outsource (to specialist die shops) if:

• Your part is a standard geometry with moderate tolerances (think brackets, simple flanges).
• You lack internal expertise for specific alloys (e.g., high-strength steel for chassis components, or aluminum for extrusions that need to fit into a catalytic converter replacement system).
• You want to avoid capital expenditure and keep fixed costs variable.

This analysis was accurate based on my experience up to late 2024. The die-making industry is evolving—additive manufacturing for die inserts is starting to change the lead time equation. So verify current pricing and lead times before making a decision. But the fundamentals haven't changed: control your core, buy your commodity. That's the rule I now live by after making (and documenting) enough mistakes to fill a small binder.