I'm experimenting with “ownership confidence”

I'm experimenting with “ownership confidence”

One thing my research keeps coming back to is that documented ownership and operational reality aren't always the same thing.

A service catalog might say:

Owner → Team A

But other evidence might say something different.

For example:

Documented ownership
Team A

Current on-call
Team A

Recent deployment
Team B

Recent engineering activity
Team B

Dependency investigation
Team B

Now what?

It would be tempting for a system to simply choose Team B and say:

“Team B owns this.”

But that could be wrong.

The evidence is conflicting.

So I'm experimenting with a simpler idea:

Ownership confidence

Instead of forcing a system to choose an owner, show:

Documented ownership → Team A

Operational evidence → Team B

Confidence → Ambiguous

And then make the reason for that uncertainty visible.

The idea is that the system shouldn't pretend to know more than the evidence supports.

It should help someone answer:

What do we know?

What evidence supports it?

Where does the evidence disagree?

What still needs to be confirmed?

This connects to something I've been researching around operational convergence.

During complex incidents, the problem isn't always finding a piece of information.

Sometimes the information already exists.

The problem is that it's distributed across different systems, teams, records, deployments, schedules and operational activity.

So I'm testing whether bringing those signals together — while explicitly showing uncertainty — can help people move toward coordinated action without creating another workflow they have to maintain.

The experiment is deliberately simple.

No AI deciding who owns something.

No automatic declaration of responsibility.

No attempt to replace existing systems of record.

Just:

Documented ownership

↓

Operational evidence

↓

Agreement / conflict

↓

Ownership confidence

↓

Next action

I'm still testing whether this is actually useful.

So I'd really value feedback from engineers, SREs, platform engineers and technical operators:

When documented ownership and operational evidence disagree, what would you actually want a system to show you before you're comfortable acting?

I'm building the experiment around that question.