How I’m Using One Real Data Project Per Week to Solve Local Problems That Actually Matter

byJP Mwangi MKH •
0

A young African data analyst at a wooden desk, focused on a laptop, papers with graphs and maps spread out, a chalkboard behind him reads “Real Data, Real Impact.” Urban Kenya backdrop with natural lighting. Professional, determined vibe.

Not Another Titanic Project

Let me be real with you. If you've spent any time dabbling in data science, you've probably seen the Titanic dataset. Or maybe played around with some random supermarket sales data, just to sharpen your pandas skills. That's all good when you're learning.

But lately, I've felt a shift.

See, I don't want to build dashboards or models that look impressive but never get used. I want to build data projects that solve real problems, serve real people, and support real decisions. Think of it like cooking a meal that someone actually eats, not just arranging it on a plate for show.

So I’m setting myself a challenge: one real, useful, locally-relevant data project per week this month.

No fluff. No filler. Just value.

The Moment It Clicked

A man in a hoodie reading a highlighted printed article at night under a desk lamp, eyebrows raised in realization. On-screen, “Aligning Data Projects with Business Goals” in bold.

This shift didn’t happen in a vacuum. It was sparked by an eye-opening piece from Pragmatic Institute titled "Aligning Data Projects With Business Goals."

Here’s the punchline that hit hardest:

"Most data projects fail because they aren’t tied to a business decision."

Let that sink in. We’re talking about talented folks pouring months into models that never influence a single action. That’s a heartbreaking waste of time, money, and brainpower.

As someone figuring out how to turn my data curiosity into meaningful, client-ready work, this was a wake-up call. I've built projects that looked cool, but didn’t change anything. From now on, the bar is higher.

If it doesn't help someone decide, it doesn't deserve to be built.

My New North Star: Useful Over Impressive

A sketched-style visual of a compass overlaying data charts, icons for decisions, pain points, and people. Minimalist but thoughtful, evokes clarity.

So what does "good" data work actually look like? Here's the mental checklist I'm now using:

  1. Start with the problem, not the dataset
    If there’s no decision on the line, I’m not interested.
  2. Focus on pain points, not just patterns
    Can this project reduce confusion, save time, or prevent a bad decision?
  3. Talk to real users early
    The goal isn’t code perfection; it’s clarity and context.
  4. Small and sharp beats big and blurry
    An MVP that helps today is better than a mega-model that never ships.
  5. Track usefulness, not just accuracy
    If no one reads it, shares it, or acts on it, it didn’t do its job.

Why One Project Per Week?

A flip calendar with each week stamped “Shipped,” showing small project notes pinned on a cork board. Simple tools — pen, paper, post-its, and a laptop in the frame.
Because clarity comes from doing, not just planning.

Shipping one small-but-valuable project each week keeps me accountable. It also helps me build a library of proof — not just of my technical skills, but of my thinking process, my storytelling, and my ability to support local decisions.

Each week, I’ll share:

  • A concise blog breakdown
  • A short LinkedIn post or thread
  • A newsletter issue that ties it all together

This is part build-in-public, part feedback loop, part mindset discipline.

The 4 Projects I’m Tackling This Month

A four-square infographic-style layout: unemployment, M-Pesa shop, small business icons, food price trends. Each square has a symbol or graph, vibrant colors, Nairobi market vibe.
I’ve chosen four projects that reflect real-world issues in Kenya (and by extension, many other places too). Here they are:

Week 1: Youth Unemployment in the County
Who’s most affected? What age group is being left behind? I’ll use local data to surface key insights.

Week 2: M-Pesa Agent Saturation Map
Is there room for one more agent in Karatina, or is the market overserved? This one’s for the would-be entrepreneurs and side hustlers.

Week 3: Kenya’s Hustle Economy — What’s Growing in 2025?
Using county business license data (where I can access it), I’ll identify which small business types are trending.

Week 4: Food Basket Price Volatility Tracker
Which basic goods are price-stable, and which ones are rollercoasters? This helps everyone from families to traders to policymakers.

Each of these will be small, scrappy, and shipped. No endless tweaking.

This Isn’t a Portfolio. It’s Proof.

Silhouetted figure planting a flag on a rocky hilltop made of scattered data files, labeled “Proof Not Portfolio.” Sunrise in background symbolizes clarity and direction.
I’m not doing this to impress hiring managers. I’m doing this to prove that data can be useful — even in short time frames, with simple tools, and under real-world constraints.

To prove that with the right mindset, you don’t need big budgets or Silicon Valley funding to build decision tools that matter.

To show that in Kenya, and across Africa, data can empower youth, entrepreneurs, local leaders, and families — not just tech insiders.

So if you’re a policymaker, a small business owner, a fellow builder, or just someone curious about how data can help you navigate uncertainty — I’m doing this with you in mind.

Let’s turn data into something that actually moves the needle.
One week at a time.

Post a Comment

Previous Post Next Post