When I was about thirteen, I decided I needed a more lucrative skill.
I’d already asked my parents what I could do to earn some extra money, and they’d given me a perfectly reasonable answer:
Clean the office.
So I did.
I got paid for it. I’m fairly sure my mother had to redo a fair amount of my work in the beginning.
Eventually, though, I was over cleaning.
I wanted to learn something else.
So I asked Mum to teach me something I could actually use to make money.
She decided to teach me bookkeeping.
This was probably a fairly unusual response to a then thirteen-year-old asking for more pocket money, but it worked out surprisingly well.
Everything is basically a database#
Mum had spent years working with business systems and databases, so when she started teaching me MYOB, she didn’t begin with a thrilling lecture about debits and credits.
Thank God.
She explained the system.
MYOB was basically a database.
FileMaker, which they also used, was basically a database.
Different front ends. Different purposes. But underneath, you’re putting information into a structured system so you can do something useful with it later.
And Mum had a phrase for this:
Everything is basically a database.
That made sense to me.
If you put good information in, you could get useful information out.
If you put rubbish in, well…
The database wasn’t going to magically fix it for you.
So I learned MYOB.
Then I learned bookkeeping.
Eventually I started doing my parents’ bookkeeping, then bookkeeping for some of their clients, and eventually I had my own little bookkeeping business while I was still at school.
And that’s where I discovered something else.
Bookkeeping is surprisingly good detective training#
A database is only useful if the information inside it bears some resemblance to what actually happened.
And sometimes figuring out what actually happened takes considerably more work than typing numbers into MYOB.
Where did this payment come from?
What was this purchase actually for?
Which customer does this relate to?
Why doesn’t this reconcile?
What the hell happened here?
Suddenly bookkeeping wasn’t just data entry.
It was detective work.
You had to establish what had actually happened, classify it correctly, put the right information in the right place, and make sure the relationships between things made sense.
Because eventually somebody was going to ask the system a question.
How much did we spend?
What do we owe?
Who owes us money?
How is the business actually doing?
And the quality of the answer depended on the quality and structure of the information we’d put in.
That idea stuck.
Then MYOB became Xero#
Later, I started working with clients who used Xero.
Different software.
Different interface.
Different buttons.
But underneath?
Basically a database.
So instead of treating Xero as an entirely new universe, I could ask:
Where does this information go?
How do I enter the same thing here?
How does this system categorise it?
What connects to what?
How do I get the reporting I need back out?
Once I understood the underlying structure, I could learn the differences.
And then the same thing kept happening.
FileMaker.
Excel.
Business systems.
Project management systems.
Different interfaces. Different terminology. Different purposes.
But I kept asking variations of the same questions.
What information is going in?
How is it being organised?
What connects to what?
And what do we need to get back out?
More data isn’t necessarily better data#
Eventually I learned another part of the lesson.
Technically, you can capture everything.
You can create another field.
And another one.
And another one.
You can record every conceivable piece of information just in case someone, somewhere, someday decides they might like to know it.
You can also spend an unreasonable amount of everyone’s life entering data nobody ever uses.
So the question changed.
Instead of:
What information can we capture?
I started asking:
What do we actually need this system to tell us?
If we need a particular report, decision, trigger, or outcome at the end, what information needs to exist at the beginning?
Start with the question you want the system to answer.
Then work backwards to the information it needs.
Because good data isn’t just accurate.
It also has to be useful.
Somewhere along the way, the idea got bigger#
Over time, “everything is basically a database” became “everything is basically a system.”
Not because I suddenly discovered some grand theory of how the world works.
I think Mum had just given me language for something my brain was already doing.
When something was unfamiliar, I looked for the familiar structure underneath it.
Once I recognised the structure, I had somewhere to put the unfamiliar bits.
A new piece of software didn’t have to be completely new.
A new industry didn’t have to be completely new.
A new project didn’t have to be completely new.
There was usually something underneath it that I already understood.
Then I could learn what was different.
I’m 37 now, and Mum and I still joke about it.
I’ll be telling her about some completely different industry, project, or system I’ve wandered into, and somehow we’ll end up back at the same conclusion.
Everything is basically a database.
Which, somewhere along the way, became:
Everything is basically a system.
And eventually, in my own slightly less elegant terminology:
It’s all the bloody same.
Originally published on LinkedIn.