What "Good Enough" Data Actually Looks Like
Perfection isn't the goal. Prevention is.
“We tried the governance route. It didn’t stick.”
I’ve heard this from more MRO leaders than I can count.
They invested in MDM. Built taxonomies. Defined standards. Trained users. Hired data stewards. Stood up governance councils.
Two years later, the data is dirty again.
So they conclude: maybe clean data isn’t realistic for operations. Maybe “good enough” means living with a certain level of mess.
That conclusion is understandable. It’s also wrong.
The Perfection Trap
Most data governance programs aim for perfection.
100% compliance. Zero duplicates. Every record classified correctly. Every description standardized.
On paper, reasonable. In practice, impossible — and counterproductive.
Here’s why: perfection requires control. Control requires friction. Friction gets routed around.
The moment your governance process slows down someone who needs a part now, you’ve lost. They’ll create the record their way. They’ll bypass the workflow. They’ll deal with the consequences later — which means you will.
Governance programs that aim for perfection achieve neither perfection nor good enough. They achieve resentment and workarounds.
What "Good Enough" Actually Means
Good enough isn’t a compromise. It’s a design philosophy.
Good enough means:
Prevention over remediation. Stop duplicates at creation, not in quarterly reviews.
Progress over perfection. 90% prevention with zero friction beats 99% prevention that everyone hates.
Defaults over enforcement. The system should guide toward the right answer, not punish the wrong one.
Speed over ceremony. If governance adds time, it will be ignored under pressure. Every time.
The 80/20 of MRO Data
Not all data problems are equal.
In most MRO operations, a small percentage of parts drive the majority of spend, failures, and criticality. These are the parts that matter — the ones where duplicates cost real money.
Good enough means focusing prevention where it counts.
A duplicate record for a $3 gasket ordered once every five years? Annoying, not catastrophic.
A duplicate record for a $5,000 pump seal on critical equipment? That’s where downtime hides.
Governance programs that treat all data equally waste effort on the trivial and miss the critical. Good enough means knowing the difference.
What This Looks Like in Practice
We’ve learned what sticks and what doesn’t.
What doesn’t stick:
- Governance councils that meet quarterly to review violations
- Training programs that teach taxonomy to people who won’t remember it
- Approval workflows that add days to part creation
- Data stewards who become bottlenecks
- Cleanup projects that fix the past but not the future
What sticks:
- Validation at the point of entry — real-time, before the record exists
- Suggestions, not mandates — “Did you mean this existing part?” beats “Request denied”
- Smart defaults — auto-standardize descriptions without asking
- Focused scope — protect the critical, tolerate the trivial
- Zero training required — if it needs training, it’s too complicated
The Outcome That Matters
Good enough isn’t about achieving a number — 95% clean, 99% compliant.
Good enough is about breaking the cycle.
You know you’ve reached good enough when:
- New duplicates stop appearing at the old rate
- Cleanup projects become smaller each year, not larger
- Users don’t complain about governance because they don’t notice it
- The reliability engineer at 2 AM can find the part they need
That’s the goal. Not perfection. Prevention. Not compliance. Flow.
Where This Leaves Us
Living inside this problem taught us that data quality isn’t a destination. It’s a system behavior.
You don’t achieve clean data. You build systems that keep data clean.
That’s why we built Ark — not as an MDM platform, not as a cleanup tool, but as a prevention layer. Lightweight. Frictionless. Present at the moment of creation.
Not perfect. Good enough — in the way that actually works.
About the Author
Author Bio: Raghu Vishwanath is Managing Partner at Bluemind Solutions and serves as CTO at KeyZane, a financial inclusion platform live in Central and West Africa. Over 30+ years across software engineering and technical leadership, he has watched the terms of specialization change — and learned that the only sustainable expertise is the willingness to build it again.