I once held a BASF‑funded PhD position researching “green” concrete-and then I left to build in industry. I didn’t leave the mindset. Materials science still informs how I design software and organizations.

Here are four ideas I carried from labs and construction sites into product work.

1) Hypothesis First, Then Pour

In research, you don’t “feel” your way to results; you form hypotheses, test, and publish the decision. The same discipline keeps product teams honest.

Decision memo skeleton (one page):

  • Context and goal
  • Options considered
  • Choice and why
  • Failure modes we accept
  • Next review date

One good memo saves ten meetings and a quarter of drift.

2) Respect the Curing Time

Concrete reaches strength over time. Product decisions do too. Rushing feedback loops creates brittle outcomes.

  • Protect true deep‑work blocks for discovery and design.
  • Timebox experiments, but don’t force conclusions before the signal is ready.
  • Freeze scope during pilot windows so results mean something.

“Fast” doesn’t mean frantic; it means stable under load.

3) Design for Durability (Not Just Launch Day)

In materials, durability is a first‑class property. In software, we often optimize for demo appeal. Shift to durability:

  • Bake compliance into workflows (not in a PDF at the end).
  • Make data lineage inspectable for audits.
  • Prefer guardrails over “tribal knowledge.”
  • Choose simplicity you can reliably operate with junior staff.

Durability is what lets you delegate without fear.

4) Sustainability as a Product Constraint, Not a Slogan

“Green” concrete isn’t about a banner; it’s about material efficiency and reuse. The software equivalent:

  • Minimize rework and variance (waste).
  • Reuse components and decisions (design systems, templates, validators).
  • Make expert knowledge legible and repeatable.
  • Optimize for the smallest honest system that delivers the unit of value.

Less entropy, more outcomes.

Closing Thought

Leaving a PhD didn’t mean leaving rigor. It meant applying it where the constraints are messy and the feedback is fast. If you want software that lasts, borrow from materials: hypothesis, curing, durability, sustainability. Your product-and your team-will hold under load.