Your most expensive data entry clerk is surprisingly good at running operations

Ever had someone in a meeting say "we can't trust what operations is telling us" while gesturing at a report with their soft, manicured hands?

Well, yeah. Because, more often than not, your best operators and supervisors have been turned into the world's most expensive and least enthusiastic data entry clerks.

The data entry tax

Here's what I saw once:

Supervisors spent up to an hour every shift filling in numbers across 3 different Excel documents. Each sheet asked for essentially the same information, just formatted and presented in a slightly different way.

The supervisors hated it. They'd rush through it at the end of shift, make mistakes because they were tired and wanted to go home, and then get called into meetings the next day to explain why the numbers didn't line up.

"Why does production show 1,247 but finance has 1,274?"

"Because I fat-fingered it at 6am after a 12-hour shift, Dave. That's why."

Meanwhile, the actual supervisor's job - you know, supervising - got squeezed into whatever time was left after playing data entry clerk.

Here's the thing about operations people: they're hired to run operations.

They're good at it. They like it. They want to be out there solving problems, keeping things moving, making sure nothing catches fire (literally or metaphorically).

When you force them to sit at a computer and fill in boxes, you get:

• Resentful supervisors who see it as pointless admin

• Rushed data entry full of errors

• "Good enough" numbers instead of accurate numbers

• Time stolen from their actual job

No wonder the numbers aren't squeaky clean.

The three sheet nightmare

Back to the 3 Excel sheets per shift.

Each sheet had been lovingly created by a different person at a different time for a different purpose. Over the years, they'd all evolved independently, like something Darwin would try to explain to you with some beans.

All the sheets were asking for the exact same physical material. It just needed to be calculated three different ways, entered three different places, and converted perfectly every single time.

What could possibly go wrong?

The supervisor had to:

1. Do the maths three different ways

2. Enter it three different places

3. Pray they didn't transpose a digit

4. Repeat for every other data point

5. Every single shift

6. While also, you know, supervising

No wonder they made mistakes. I'd make mistakes too.

What actually worked

The fix wasn't rocket science, but it required actually respecting the supervisor's time:

One sheet to rule them all

Consolidated all data entry into a single Excel form. Enter the number once, in one format. Done.

Automate everything that can be automated

If the data exists somewhere - anywhere - copy it across or pull it automatically. Don't make humans re-type things computers already know.

This cut manual data entry by about 80%. Supervisors only entered what genuinely couldn't be automated: delays, incidents, notes about why things went sideways.

Data entry time went from an hour to about 10 minutes.

Enforce structure, not creativity

Strict naming schemas. Dropdown lists. Data validation. Make it physically impossible to enter garbage data even if you're tired, rushed, and daydreaming about going home.

Equipment names had to match the master list. No variations. No nicknames. No "that truck" or "the dodgy one."

If it's not in the dropdown, it doesn't exist. Fight me. (Actually, please don't, I'm just a little data guy).

Connect it to everything downstream

Built ETL flows that pulled from this single source and fed all the downstream reports automatically.

Production dashboard? Auto-updated. Finance reconciliation? Auto-updated. Executive summary? Auto-updated. Rob's special report that only Rob looks at? Also auto-updated.

The only way to fix a number in any report was to go back to the source sheet and fix it there. One place. One truth. No more "which version is correct?"

The result

Supervisors went from spending an hour on data entry to 10 minutes.

They stopped making mistakes because they weren't rushing and they weren't converting the same number into three different units while mentally composing their resignation letter.

Downstream reporting became reliable because it was all pulling from one clean source instead of three slightly different interpretations of reality.

And most importantly: supervisors went back to doing their actual job, which was supervising operations, not cosplaying as reluctant data entry clerks.

Why this matters

That data triangle everyone talks about (granular data > aggregated reports > executive dashboard) breaks at the bottom not because the concept is wrong, but because we're asking the wrong people to feed it.

Operations people aren't bad at data entry because they're careless or incompetent.

They're bad at it because:

1. It's not their job

2. They don't want to do it

3. You're taking them away from work they're actually good at

4. They're doing it at the end of a long shift when their brain has left the building

If your supervisors are spending significant time every shift filling in boxes on a computer, you have two problems:

1. Your data is probably unreliable (because they're rushing through it to get back to real work)

2. Your operations are probably suffering (because your best people aren't supervising, they're doing admin)

It's a lose-lose situation dressed up as "data governance."

What to do instead

1. Automate everything you possibly can

If the data exists somewhere - anywhere - in a system, pull it automatically. Don't make humans re-type information that computers already have. This isn't the 1980s.

2. Consolidate data entry into one place

One sheet. One time. Not three sheets asking for the same information in slightly different formats because "our department needs it this way."

Your department can do a conversion. Supervisors can't do three conversions without making mistakes.

3. Enforce structure ruthlessly

Make it impossible to enter bad data. Dropdowns, validation, naming schemas. If they can't accidentally break it, they won't deliberately break it either.

4. Make manual entry as fast as possible

If a supervisor can't complete data entry in 10-15 minutes max, you're either asking too much or your process is broken. Probably both.

5. Respect their actual job

These people are operations experts, not data entry clerks. Treat their time accordingly. Every minute they spend typing is a minute they're not supervising.

The real test

Here's how you know if you've got this right:

Ask your supervisors how long they spend on data entry each shift.

If the answer is more than 15 minutes, you're wasting their time and getting unreliable data in return.

If they sigh heavily before answering, you're definitely wasting their time.

If they immediately launch into a detailed rant about how painful it is, you've got a serious problem.

If they look around nervously before answering because they're worried about getting in trouble for admitting how long it actually takes, you've got an even bigger problem.

Your best operators should be out there operating, not sitting at computers wishing they were literally anywhere else.

Fix the data capture layer by respecting the people doing the capturing. Automate what you can. Simplify what you can't. Get them back to their actual job.

The data gets better. The operations get better. Everyone wins.

Except maybe the person who built those 3 Excel sheets. They're going to be a bit upset when you delete two of them.

But that's a sacrifice I'm willing to make.