Data Plumbing: Getting the Basics Right Before You Dream Big

Let's talk about plumbing. Not the kind under your sink (though that's important too), but the kind that moves data through your business. Because here's the thing: no matter how fancy your dashboards look, how clever your AI sounds, or how many buzzwords you can fit into a strategy deck — if your data plumbing is leaky, clogged, or just plain broken, all that value? It evaporates.

Would You Want “Innovative” Plumbing?

Imagine this: you walk into your bathroom and see a toilet with a touchscreen, a voice-activated shower, and a mirror that gives you the weather. Sounds cool, right? But then you flush the toilet and it backs up. The shower sprays cold water. And the mirror? It's fogged up because the extractor fan doesn't work.

Would you call that innovative? Or just a mess? That's what happens when businesses chase innovation — AI, automation, predictive analytics — without fixing the basics. Without solid, reliable, secure data plumbing.

The Data Pyramid: From Raw to Remarkable

Let's break it down using the extended data pyramid. Think of it like building a house:

1. Data — raw bricks. Useless on their own.
2. Information — bricks stacked into walls. Now we're getting somewhere.
3. Knowledge — rooms with purpose. You know what's where.
4. Wisdom — a home that works for you. You understand how to live in it.
5. Action — you start using it. Cooking, sleeping, relaxing.
6. Automation — the lights turn on when you walk in. Nice.
7. Prediction — the thermostat knows when you'll be cold.
8. AI — the house learns your habits and adapts.

But here's the catch: if the plumbing isn't right at level 1, everything above it is at risk. You can't automate a cold shower. You can't predict a flood if your sensors are broken. And you definitely can't trust AI if it's learning from bad data.

What Is Data Plumbing, Really?

Data plumbing is the unglamorous, behind-the-scenes work that makes everything else possible. It's:

• Making sure data flows from source to destination without leaks.
• Ensuring it's clean, consistent and complete.
• Validating that what goes in is what comes out.
• Monitoring for blockages, breaks, or weird smells (metaphorically speaking).

It's not sexy. It's not flashy. But it's essential.

The Hidden Cost of Bad Plumbing

Let's say your sales dashboard shows a 20% drop in revenue. Panic! You call a meeting. People scramble. Strategies shift. But then someone checks the pipeline and finds out the data feed from your CRM failed last night. The drop? Not real. What would be worse is if the 20% drop is real, but the data can't tell you where that drop is and what is causing it — now you've still got the disaster and a complete inability to do anything about it.

That's the cost of bad plumbing. Not just money, but time, trust and morale. And it isn't just about dashboards. Bad data plumbing can:

• Break customer experiences (an existing customer buys a new product, but you don't recognise them).
• Mislead decision-makers (bad forecasts, wrong KPIs, no ability to triage and diagnose problems).
• Undermine analytics and AI models (garbage in, garbage out).

Secure, Reliable, and Boring (In a Good Way)

Good plumbing is boring. It just works. You don't think about it. That's the goal.

• Secure: no leaks. No breaches. No surprises.
• Reliable: it works every time, not just on Tuesdays.
• Monitored: you know when something's wrong before it becomes a crisis.

And just like real plumbing, it needs regular maintenance. You don't wait for a flood to check your pipes.

Enter Data Contracts: The Plumber's Blueprint

One of the best tools in modern data plumbing is the data contract. Think of it as a blueprint between the people who create data (engineers, product teams) and the people who use it (analysts, data scientists, AI systems). It says:

• “This is what the data will look like.”
• “This is how often it will arrive.”
• “This is what it means.”

And if something changes? Everyone knows. No surprises. No silent breakages. It's like agreeing on the size of the pipes before you start pouring water through them.

Fixing the Leaks: A Practical Approach

You don't need to rip everything out and start over. Start small:

1. Map your pipelines — know where your data comes from and where it goes.
2. Add monitoring — set up alerts for failures, delays or weird values.
3. Define contracts — start with your most critical data flows.
4. Clean as you go — don't wait for a big “data quality project”. Fix issues when you find them.
5. Talk to each other — data isn't just a tech problem. It's a team sport.

From Plumbing to Prediction

Once your plumbing is solid, the rest of the pyramid becomes possible:

• Information becomes trustworthy.
• Knowledge becomes actionable.
• Wisdom becomes strategic.
• Action becomes confident.
• Automation becomes safe.
• Prediction becomes accurate.
• AI becomes valuable.

But skip the plumbing? And you're building a smart home on a swamp.

Final Thoughts: Back to Basics

In a world obsessed with innovation, it's easy to overlook the basics. But the basics are what everything else stands on. So before you chase the next big thing — before you invest in AI, automation, or the latest analytics tool — ask yourself:

Is our data plumbing solid?

Because if it's not, you're not building. You're patching. And no one wants to live in a house with “innovative” plumbing.