Why Co-Development Still Fails

I’ve watched co-development fail in every way it possibly can.

• Quietly.

• Expensively.

• Politely.

Sometimes the partner gets blamed. Sometimes the geography. Sometimes “communication.”

After 24 years and more than 400 shipped titles, I’ve learned something uncomfortable:

When co-development fails, it’s almost never for the reasons leadership says out loud.

I first wrote a version of this article in 2015, when “outsourcing” was still treated as a transactional necessity, something you bolted onto a production when things were already on fire.

A decade later, co-development is no longer optional. It’s how modern games are built.

What hasn’t changed is why it fails.

What has changed is that we now understand the difference between outsourcing and orchestration. Between renting hands and building outcomes. Between vendor management and shared accountability.

Co-development doesn’t fail because of geography.

It fails because of misalignment, weak systems, and leadership avoidance.

Here are five reasons it still breaks, and what we’ve learned the hard way.

You can’t hit a moving target.

And yet most co-development engagements begin with documents that look complete until real production starts.

Specs drift because:

• Decisions aren’t owned

• Trade-offs aren’t surfaced early

• Reality is postponed in favor of optimism

For years, I believed this was a production discipline problem. Better documentation. Tighter control. More reviews.

I was wrong.

Specs don’t drift because teams are careless. They drift because leadership avoids making hard calls visible.

At Streamline, we learned that specs can’t be treated as static documents. They are living commitments, tied to owners, milestones, risk, and cost.

When something changes, it must be explicit:

• What moved

• Why

• Who approved it

• What it costs

No ambiguity. No silent scope bleed.

Principle:

Clarity is cheaper than heroics. Every time.

Co-development collapses fastest when procurement logic overrides production reality.

If a quote feels too good to be true, it usually is.

And when you squeeze a partner below sustainability, you don’t get savings, you get:

• Rework

• Attrition

• Delay

I’ve seen this pattern repeat across studios of every size. The spreadsheet says you saved money. The schedule and team morale say otherwise.

What matters isn’t day rates. It’s the total cost of outcome:

• Iteration cycles

• Integration friction

• QA loops

• Producer load

• Opportunity cost of delays

Cost control without delivery control isn’t discipline. It’s gambling with better spreadsheets.

Principle:

If you don’t price for reality, reality invoices you later.

Most co-development partners have seen more broken pipelines, late pivots, and last-minute saves than internal teams ever will.

Yet studios routinely ignore that experience.

I’ve been in rooms where the people closest to the work saw the problem early and didn’t feel they had permission to say it. Weeks later, leadership discovered the same issue at ten times the cost.

That’s not a partner problem. That’s a trust problem. When something starts to go sideways, the most valuable question leadership can ask is simple:

“What are you seeing that we’re missing?”

That question alone has saved productions.

Principle:

If you don’t trust your co-development partner’s judgment, you chose the wrong partner.

Co-development doesn’t scale on goodwill.

It scales on systems.

Email chains, slide decks, and status meetings do not survive:

• Time zones

• Iteration velocity

• Modern content volume

Culture matters, but infrastructure decides whether culture survives contact with reality.

We rebuilt our operating layer around a single source of truth, one that shows work as it exists, not as it’s reported:

• In-engine visibility

• QA-first routing

• Live reporting without theater

• Signal for leadership, not noise

That’s how distributed teams operate without drowning in coordination debt.

Principle:

If your system can’t show reality in real time, reality will show up late and expensive.

This is the quiet killer.

Co-development fails when it’s managed by:

• Inexperienced producers learning under fire

• MBAs with no shipping scars

• Leaders without mandate or authority

For too long, the industry treated co-development as a side function, something junior, something to delegate, something to buffer away from “real” leadership.

That belief is costly.

Co-development leadership isn’t a title. It’s earned through pressure. Through decisions made close to the work. Through escalation paths that are fast and real.

Middle layers that exist to buffer discomfort don’t protect teams, they delay truth.

Principle:

Co-development is not a support function. It’s a core business capability.

Co-development isn’t about outsourcing anymore.

It’s about orchestration.

• Global execution with earned trust

• Experience over headcount

• Systems over promises

• Delivery over theater

I’ve seen what happens when leadership engages honestly and when it hides behind distance until the bill arrives.

Only one of those paths ships and everyone knows which one they’re on.

Photos by the author. Taken in Bangsar, 2019.