Data Engineers Don’t Burn Out from Work They Burn Out from Pointless Work

Learn how Co-op’s Head of Data Engineering is redesigning delivery to reduce burnout, cut cloud costs, and build trust in your data.

 

Rethinking the Way We Scale Data Teams

Data leaders today aren’t struggling with a lack of tools, they’re drowning in dashboards, always-on pipelines, and the relentless expectation to deliver “business value” at speed. The result? Teams are overworked, under-appreciated, and burning out, not because they’re lazy or incapable, but because so much of the work they’re doing doesn’t matter.

In this episode of the Data Matas podcast, Nik Walker, Head of Data Engineering at Co-op, shares how he’s building resilient, human-first data teams in one of the UK’s most complex legacy organisations. It’s a conversation about slowing down to speed up, leading without ego, and refusing to confuse velocity with value.

You’ll learn how to:

• Innovate where it counts by choosing real-time, batch, or hybrid delivery based on actual need
• Build trust in your data before you go near AI adoption
• Design structured, psychologically safe teams without performative leadership
• Prioritise discovery over delivery, and avoid burning out your best people
• Navigate modernisation without ripping out every legacy system

This article is based on the Data Matas podcast, which delivers actionable insights you can take away and apply to your business.

 

Meet Nik Walker

Head of Data Engineering, Co-op

Nik leads a team of engineers transforming Co-op’s data infrastructure, spanning retail, insurance, funeral care, and more. But what makes Nik stand out isn’t just his technical depth, it’s his relentless focus on people.

“Leadership isn’t about putting your arm around everyone. If every conversation is safe and every outcome is clear, that’s how you look after people.”

He’s a self-described “human-centric” leader who’s as comfortable talking about pipelines as he is about psychological safety and neurodiversity. With no formal academic background and a personal history shaped by ADHD, Nik brings a grounded, unfiltered approach to leadership that cuts through the noise.

 

The Real Problem Isn’t Technical Debt

It’s Pointless Work

Let’s be honest—most data teams aren’t failing because the tech isn’t good enough. They’re failing because they’re stuck in cycles of work that doesn’t move the needle. Endless dashboards. Ad hoc requests. Real-time syncs no one asked for.

“People don’t burn out from work. They burn out from pointless work.”

Nik’s team actively filters out low-value work. They use impact sizing to decide what gets done, what gets parked, and, crucially, what gets a hard “no.” That clarity is a survival strategy, not a luxury.

 

How to Put This Into Practice:

1. Audit your workload – Map what your team is doing weekly. Tag tasks as BAU, strategic, or reactive.
2. Create an impact-sizing model – Score projects on business value, urgency, and technical effort.
3. Give people permission to say no – Make it a leadership responsibility to defend your team’s time.

Expected Result: More time spent on things that actually deliver value—and a team that isn’t exhausted.

 

Real-Time Isn’t Always the Right Time

A major source of wasted effort (and budget) is overengineering. Nik calls out a common trap: syncing everything, all the time, just because you can.

“You don’t need real-time data. You need the right-time data.”

By asking “what decision is this data supporting?”, Co-op has focused on incremental models, reduced compute load, and avoided unnecessary data processing.

 

How to Put This Into Practice:

• Map syncs to business questions – Does the decision need real-time input? If not, don’t sync in real-time.
• Use delta loads instead of full loads – Nik’s team found huge savings in incremental processing.
• Schedule syncs based on usage – Use logs to spot when data is actually accessed and align your processing to that.

Expected Result: Lower cloud costs, less unnecessary processing, and fewer team hours wasted on firefighting failed jobs.

 

Build Trust Before You Build AI

Nik is blunt about the risks of rushing into AI without a solid foundation. If the team doesn’t trust the output, the initiative dies before it starts.

“If I don’t trust the numbers, what’s the C-suite going to think?”

While LLMs and GenAI dominate the hype cycle, Co-op’s focus is on governance, quality, and observability. No AI gets near production until the maths behind the scenes is trustworthy.

 

How to Put This Into Practice:

• Treat trust as a deliverable – Log where numbers come from. Show your working.
• Don’t assume “the model is right” is good enough – Build explainer tools or simple audit paths into your data flows.
• Run in shadow mode – Let AI systems operate in the background before surfacing them to users.

📊 Expected Result: AI that augments, not undermines—your data strategy.

 

Discovery Is Not Optional

Nik’s team never builds without first doing discovery. That means understanding the need, mapping data flows, and designing with intent. It’s not slow, it’s smart.

“We don’t get given the time or respect to do discovery properly. So we end up rushing—and burning out.”

Too many teams jump straight to solutions. But when discovery gets skipped, you end up solving the wrong problem with the wrong tech. Every. Time.

 

How to Put This Into Practice:

• Protect time for scoping – Block out discovery sprints before build cycles.
• Ask the problem twice – If someone requests a dashboard, ask what decision it’s enabling.
• Make discovery visible – Share your problem definitions, diagrams, and learnings before you ship anything.

Expected Result: Fewer failed projects, stronger stakeholder buy-in, and more resilient systems.

 

Care Is a System, Not a Slogan

Nik is passionate about creating psychologically safe environments, but not in the fluffy, performative way. For him, caring means systems, structure, and clarity.

“If every conversation is safe, if every outcome is clear—that’s how you take care of people.”

He’s designed a framework that lets his team focus, feel supported, and build without fear. And it’s working.

 

How to Put This Into Practice:

1. Systematise support – Document what meetings are for, what decisions are expected, and who owns what.
2. Normalise neurodiversity – 60% of data professionals are neurodivergent. Assume difference is the norm.
3. Reward calm, not chaos – Promote people who create stability, not just speed.

Expected Result: Teams that scale sustainably, without burning out the people who keep them running.

 

Start Here: Your Implementation Plan

If you’re serious about building a healthier, more effective data team, start with the sequencing:

1. Week 1–2: Audit your team’s time and impact
2. Week 3: Run a “stop doing” workshop and implement impact scoring
3. Week 4–6: Review sync schedules, convert full loads to incremental where possible
4. Month 2: Introduce discovery sprints on all new data initiatives
5. Ongoing: Make psychological safety and clarity part of your team rituals
   

📊 Metrics to Track:

• % of time spent on reactive work
• Cost of data processing (pre vs. post optimisation)
• Team satisfaction and burnout risk (via internal pulse checks)

Real-World Results from Co-op

Co-op is already seeing results. By switching from full loads to delta processing and questioning the need for real-time syncs, they’ve cut compute costs. They’ve also introduced a stronger governance layer, so teams now build with clarity and purpose, not panic.

And maybe most importantly? Engineers aren’t dreading their Mondays.

 

Where You Go From Here

The path to scalable, trustworthy data systems doesn’t start with new tools. It starts with better choices, about where your team’s energy goes and what outcomes you’re building for.

Remember these takeaways:

• Stop chasing speed and prioritise meaningful work
• Optimise syncs based on actual business need
• Build trust in your data before chasing AI trends
• Protect your team’s energy by building systems that care

Listen to the full conversation with Nik Walker on the Data Matas podcast, it’s packed with practical insights you can apply to your team right now.

Dig Deeper

🎙️ Listen to the full episode
 🔗 Connect with Nik Walker on LinkedIn
 🌍 Visit Matatika’s Website
 📺 Subscribe to Data Matas on YouTube

 

#Blog