You Don't Have a Data Problem. You Have a Process Problem. (Copy)
Everyone says the same thing.
"Our data's a mess."
"Our spreadsheets are unreliable."
"We can't trust the numbers."
So they think the solution is obvious: fix the data.
Clean it up. Validate it. Build better checks. Hire someone to babysit it.
And yes, the data is messy.
But that's not the actual problem.
Here's What's Really Happening
Let's say you've got an inventory spreadsheet.
It's been running for three years. Sarah built it originally. Now Mark uses it. Sometimes Tom adds data to it. Nobody really owns it.
The data's full of inconsistencies:
Different date formats
Typos in product codes
Duplicate entries
Missing fields
Figures that don't reconcile to anything else
Your instinct: the data's broken, so we need better data.
But here's the thing.
The data isn't broken because the data's broken.
The data's broken because nobody owns the process that creates it.
Sarah built the spreadsheet to solve her problem in 2021. It worked for her.
Now four other people use it for different reasons, and nobody updated the rules.
There's no standard for how to enter data. No documentation. No checks. No accountability for quality.
When Sarah left, the process left with her.
And the data got worse.
The Data Problem Is a Symptom
This is the part people miss.
You can spend weeks cleaning up data:
You can build validation rules
Add alerts
Create flags
Hire a data analyst to police it
And six months later, you're back where you started.
Because you fixed the symptom, not the cause.
The cause is process.
Who enters the data? When? In what format? What happens if it's wrong? Who checks? Who's responsible if something breaks?
These questions don't get asked. So the process drifts. Everyone does it slightly differently. Standards disappear. And the data reflects that chaos.
In manufacturing and operation-heavy businesses, this is brutal.
Your production numbers feed into costing.
Your costing feeds into pricing.
Your pricing feeds into margins.
If the production data is inconsistent—because nobody owns the entry process—then your costing is wrong. Your pricing is wrong. Your margins are wrong.
You're making business decisions on data you can't actually trust.
And you're blaming the data instead of the process.
Why This Matters
Here's what happens when you focus on fixing the data instead of the process.
You get a temporary improvement.
For about three weeks, the data looks better. It's clean. It's consistent.
Then the improvement decays.
New data comes in without the same care. Edge cases emerge that nobody anticipated. People get busy and take shortcuts. The process slips back because there's no structure holding it up.
It's like cleaning a messy room without changing the habits that made it messy in the first place.
You'll spend the rest of your life cleaning.
When you fix the process, the data stays clean.
Because the process itself maintains the standard.
What a Real Process Looks Like
A process owns three things.
First, the rules.
How does data get entered?
What format?
What values are acceptable?
What's the definition of each field?
These aren't optional.
They're documented and they're enforced.
Second, the responsibility.
Who enters the data?
When?
Who checks it?
Who fixes it if it's wrong?
There's no "data gets entered by whoever" situation.
There's a person or a team. There's accountability.
Third, the feedback loop.
What happens when something goes wrong?
Does anyone notice?
Does anyone fix it?
Or does bad data just cascade downstream, making wrong decisions look right?
When these three things are in place, data quality becomes a consequence of the process, not a thing you have to police constantly.
You're not hiring someone to babysit data.
You're building a system that makes good data the path of least resistance.
The Real Cost of Ignoring This
Let's say you skip this.
You've got messy data. You know it. Everyone knows it. But you're used to it, so you muddle through.
What does that actually cost?
Reports take forever because you have to manually verify and fix data before you can report it.
Decisions get made on incomplete or unreliable information because the data's not trustworthy, so people either ignore it or guess instead.
Your team spends time they don't have verifying numbers instead of using them to solve problems.
In manufacturing, a costing error from bad production data means you might be under-quoting jobs. One £20k job with wrong costing and suddenly your margin's gone.
How many of those does it take before you notice?
And when you do notice, how long does it take to find the root cause?
If the process isn't documented, you're tracing through chaos.
How to Actually Fix This
The fix doesn't start with data.
It starts with process design.
You ask:
How should this data be entered?
By who?
When?
In what format?
What checks should happen automatically?
What checks should happen manually?
Who owns the quality of this data?
Once you've got answers, you build the process.
Maybe that's a form instead of free-text entry
Maybe that's a checklist before data gets accepted
Maybe that's a weekly review where someone validates the week's entries
The structure depends on your business.
But the principle is the same: the process protects the data quality, not the other way around.
Then you document it.
Not as a novel. As a clear, short set of rules that anyone new can follow.
Then you enforce it.
For real. Not "this is the process but if you're busy you can skip it." Either the process is important or it isn't.
If it's important, you follow it.
If it's not, you admit it and stop wasting time.
This Changes Everything
Once you've got a solid process, the data takes care of itself.
You don't need a full-time babysitter.
You don't need to clean data constantly.
You don't need to spend hours every week verifying numbers that should already be right.
Your team uses the data instead of fixing it.
And your decisions are based on numbers you actually trust.
For manufacturing especially, this is the difference between reactive and strategic.
Reactive is: "Our numbers are messy, so we can't trust them, so we guess."
Strategic is: "Our process is solid, so our data's reliable, so we can see actual trends and inefficiencies."
One feels safer. The other makes money.
Where to Start
You don't need to fix everything at once.
Pick your most painful data problem.
Inventory. Costing. Production hours. Whatever costs you the most time or creates the most friction.
Ask:
What is the process that creates this data?
Who enters it?
When?
In what format?
What checks happen?
Who's accountable?
If you can't answer those questions clearly, that's your starting point.
Document what actually happens right now.
Not what should happen. What actually happens.
Then ask:
What's broken about this process?
Where do people take shortcuts?
Where do they guess?
Where does data get lost or duplicated?
Most of the time, you'll see the problem immediately.
Then you redesign.
Not from scratch. Just better.
Clearer rules. Clear accountability. Clear checks.
And you document it so the next person doesn't have to reverse-engineer what you meant.
The Outcome
Once you've got a real process, data quality becomes boring.
It stops being a crisis.
It stops being something you think about constantly.
It just works.
And your team stops spending half their time on data maintenance and starts spending it on what actually matters.
That's the shift.
That's when you go from reactive to strategic.
If you’d like some help, let’s book a free exploration call.
Office Mango helps manufacturing and distribution businesses build operational systems that work. That usually means Excel. Sometimes it means custom tools. Always, it means processes that are documented, owned, and actually followed. We offer three ways to work together: one-off Reporting Automation projects, Bespoke Solutions for bigger challenges, and Retainer arrangements for ongoing partnership.
