Sometimes teams keep applying Lean tools, yet performance barely moves. That’s a sign the problem isn’t the process — it’s the architecture behind the process.

Here are the classic indicators:

1️⃣ Improvements deliver smaller and smaller benefits.
This means you’ve reached the ceiling of the current design.

2️⃣ The real constraint lies outside the area being improved.
Handoffs, approvals, technology, or data issues block progress.

3️⃣ The process is built on assumptions from another era.
If the world around the process changed, its structure may no longer make sense.

4️⃣ Speed matters more than efficiency.
If velocity is critical, and the process can’t accelerate, reinvention becomes necessary.

5️⃣ Scaling exposes weaknesses rather than amplifying strengths.
A sign that the design was never meant to support growth.

Improvement is essential…
But knowing when to stop improving and start rethinking is an advanced leadership capability.

Engagement Question:
👉 Which of these signals have you personally experienced?

I’ve lived through signal #5 (“Scaling exposes weaknesses rather than amplifies strengths”) and it almost killed us.

Aerospace Tier-1, European site, machining large monolithic aluminium wing-ribs for the A350-1000.

For 12 years we had the most beautiful, textbook Lean cell you’ve ever seen:

• One-piece flow

• 7-axis machines with auto-tool change

• OEE 89–91 %

• 38-second takt

• Zero inventory between ops

• Gold-level 5S

• Kaizen board full of green stickers

Everyone came to benchmark us. We were the poster child.

Then in 2023 Airbus asked us to triple the shipset rate for the -1000 (from 4 to 12 shipsets/month) on the exact same footprint and headcount.

We said “no problem, we’re Lean!” and started the usual improvement circus.

Result after 9 months of heroic effort:

• OEE dropped to 62 %

• Overtime went from 4 % to 38 %

• Quality escapes tripled

• Lead time went from 11 days → 34 days

• People were crying in the canteen

Scaling didn’t amplify our strengths — it exposed that our entire architecture was secretly designed for low-rate, high-margin production, not high-rate anything.

The hidden 1990s assumptions that broke us:

1. Tool-life was calculated for 4 shipsets/month — at 12 shipsets we were changing inserts every 40 minutes instead of every 6 hours.

2. Fixtures were designed for manual load/unload — at triple rate the operators couldn’t keep up physically.

3. CNC programs were 48 000 lines long with zero modularity — every new rib length needed 3 weeks of re-posting.

4. Quality inspection was 100 % manual with calipers and templates — we simply didn’t have enough inspectors on the continent.

We had spent a decade making the wrong system go faster instead of building a system that could scale.

The moment leadership finally admitted we had to stop improving and start reinventing:

The plant manager walked past the cell at 2 a.m., saw 42 machined ribs waiting for inspection because we only had 9 qualified inspectors in the building, looked at me and said:

“If we do one more kaizen on this cell I’m going to burn the suggestion box.”

Two weeks later we killed the sacred cell.

We threw away €9 million worth of fixtures, bought two new 5-axis machines with robotic loading, rewrote every program in modular sub-routines, moved to 100 % on-machine probing, and turned the whole thing into a flexible transfer line.

18 months later we’re shipping 13 shipsets/month with lower headcount and OEE back above 90 %.

The advanced leadership capability is exactly this:

Knowing when to stop polishing the old cathedral and start demolishing it to build an airport.

Most leaders never develop it because demolition looks like failure on their yearly scorecard.

The ones who do develop it usually have to watch their people cry at 2 a.m. first.