Thumbnail image

Not Shipping Stuff? XPs Secret Weapon: Collective Ownership Beats Silos Every Time

🚀 Specialization => Silos (And Why That’s a Problem)

Many will read this and think they already do it. But there is always more to do especially as scale always attempts to reinvent this problem in new ways. What was once the solution to last weeks problem is now the cause of todays problem.

This is a process that should continue left all the way up the chain.

You hired a “Deployment Wizard.” Then they needed a “CI/CD Specialist.” Then a “Cloud Infrastructure Guru.” Now you’ve got a whole department that “owns” deployments… and suddenly, no one else can ship code without them.

This happens with product, delivery, qa AI, Data etc. ‘New’ roles are especially vulnerable due to lack of understanding.

As roles specialise everyrole thinks they are a snowflake and we start to discover that everybody in each specialism is saying the same thing in different langauges, but not translating accross. This leads to echo chambers of dev, product, delivery and service all saying the same things and thinking they owb ‘agile’ or whatever it is ‘value delivery’, ‘service orientation’ etc.

e.g. It is hard for many people to understand that data engineering is a role that has been done since the beginning of computing. Or that before product somebody still did product, or that before delivery we still managed to deliver, or that before scrum the job of a scrum master was already being done.

Oops.

We create siloes because we organise around how we hire but we dont have to.

Why organizing by skill backfires:

  • Bottlenecks form (“Sorry, only Bob can fix the pipeline”).
  • Knowledge hoarding (Tribal knowledge = single points of failure).
  • Blame games (“That’s not my job” becomes your team’s motto).
  • Throwing over the fence (no single point of accountability == nobody has accountability => nobody responsible either)

💡 XP’s Fix: Collective Ownership (Of Code, Problems… Even People!)

Extreme Programming says: “If the team owns the code, the team owns the problems.”

No more:

  • “That’s backend’s problem.”
  • “Talk to QA about it.”
  • “The DevOps team will handle it.”

Instead: “We all swim together.”

🔧 How XP Practices Make It Work

Pairing + Ensemble → No More “Hero Culture”

  • Junior + senior devs work together.
  • Frontend + backend share keyboards.
  • Engineering, product and delivery work together.

where there were once walls there are now valued and invited perspectives

Result: Knowledge spreads like gossip.

Shared Code → No Sacred Cows

Anyone can fix anything (yes, even that “scary” legacy module).

Result: Fewer bottlenecks, fewer “urgent” escalations.

Cross-Role Teams → No More “That’s Not My Job”

Devs learn deployment. Testers write automation. Product lives why estimation is a farce Decisions get made on the spot by the people who own them

Result: Faster cycles, happier teams.

📈 Real-World Win: 1000% Faster Delivery (Yes, Really)

A mid-sized SaaS team was drowning in handoffs: In 6 months nothing had been deployed, built or even agreed in specification. 7 weeks later deployed to a production environment for integration with client.

Product producted, Devs coded → QA tested → Ops deployed.

Each step had wait times, each step had internal failure rates and loop back failures.

They switched to:

  • Ensemble devs + QA + product during development, platform brought in as valued enablers when appropiate .
  • Rotating “deployment duty” (so everyone learned it).
  • No “code ownership” (anyone could fix anything).

Outcome?

  • Work started to flow
  • Fewer midnight fires (“Because everyone understood the system”). -Happier teams (“Less us vs. them”).
  • Highest performing metrics in organisation

## 🤔 “But Won’t Chaos Ensue?” (Nope.) Myth: “If everyone owns it, no one owns it.” Reality: “If everyone owns it, everyone cares.”

How to avoid anarchy:

  • Daily standups (Quick syncs, not status reports).
  • Automated tests (Safety net for changes).
  • Mentorship (Pairing ensures skills transfer).

🛠️ Alternative Structures That Don’t Suck

If “collective ownership” sounds too radical, try:

Feature Teams

Cross-functional squads (dev, QA, ops) own outcomes, not tasks.

Rotation Systems

“Ops week” for devs, “Coding day” for testers.

Guilds Over Silos

Communities of practice (e.g., “Testing Guild”)—but work stays in teams.

🎯 Why Tech Leaders Should Care

  • Faster delivery (Fewer handoffs = fewer delays).
  • Lower bus factor (No more “If Bob quits, we’re screwed”).
  • Better morale (Teams > empires).

🔥 One-Liner to Steal for Your Next Retro

“Specialization is for ants. We’re building software as a community.”