At DOES15, we presented the work we'd done at Salesforce to take their SRE teams to the "blameless cloud." We worked with various roles in the SRE teams so they could start asking the right questions about failure, and through the postmortem and retrospective process, begin to make lasting changes in how Salesforce worked with and remediated identified failures.

But DevOps espouses less siloed thinking and more shared responsibilities, so we found postmortems within the SRE organization weren't enough. As Salesforce was moving toward a model of "service ownership," teams along the entire software delivery value stream needed to start to understand their roadblocks to remediation and what aspects of the complex system they worked in were impeding their ability to "own their service."

We'll discuss the second phase of our work in helping these operations _and product_ teams gain a deeper understanding of service ownership, and why just "DevOps'ing it up" wasn't quite enough on its own to help. plus we'll introduce an expanded model from last year's talk that incorporates human factors and complexity theory. These additions helped prime the teams to more effectively grapple with the challenges facing them on the road to true service ownership.