Skip to content

No SRE Early in the Delivery Pipeline

The Lesson

Site Reliability Engineering (SRE) should be engaged early in the project lifecycle, not just before go-live.

Production readiness is not something that can be added at the end of a delivery project. It must be considered throughout design, implementation and testing. When SRE is only engaged during the onboarding phase, there is often insufficient time to address gaps in operability, observability, resilience and supportability before responsibility transfers from the delivery team to SRE.

The result is a platform that may function correctly from a development perspective, but is difficult, expensive and risky to operate in production.

The Problem

A common delivery lifecycle follows a pattern similar to the one below:

Proposal
    │
Discovery
    │
Sale
    │
Delivery
    │
SRE Assessment
    │
Handover to SRE
    │
Delivery Team Rolls Off
    │
SRE Operates the Service

The issue is that SRE involvement frequently begins only during the assessment and onboarding phase, when the project is approaching go-live.

At this stage, SRE often identifies deficiencies in areas such as:

  • Observability and monitoring
  • Alerting strategy
  • Logging quality and retention
  • Operational documentation
  • Disaster recovery and backup validation
  • Security hardening
  • Production deployment processes
  • Testing of operational scenarios
  • Capacity planning
  • Runbooks and operational procedures
  • Support model definition
  • Service ownership and escalation paths

These are rarely small changes. Many require architectural decisions, application changes or infrastructure redesign. By the time they are discovered:

  • Delivery budgets are largely exhausted.
  • Development resources are preparing to leave the project.
  • Delivery milestones are already committed.
  • There is little appetite to delay go-live.
  • There is insufficient time to properly remediate the issues.

As a result, SRE inherits a platform that does not meet an acceptable operational standard.

This creates several long-term problems.

Increased Operational Overhead

Engineers spend more time performing manual investigations, repetitive operational tasks and responding to avoidable incidents because the platform lacks the tooling and automation needed for effective support.

Reduced Reliability

Insufficient monitoring, resilience testing and operational validation increase the likelihood of production incidents and make failures harder to detect and recover from.

Longer Incident Resolution

Poor logging, missing telemetry and incomplete documentation make troubleshooting slower and increase Mean Time to Resolution (MTTR).

Increased Operational Risk

Weak disaster recovery processes, incomplete operational procedures or insufficient security controls leave the service exposed to avoidable operational risks throughout its lifetime.

Higher Cost of Ownership

Problems that could have been addressed during delivery become significantly more expensive once the service is live. Changes require operational planning, change management and production deployments, often involving multiple teams.

Reduced Customer Satisfaction

Production is where customers experience the service. Poor operational readiness results in reduced availability, degraded performance and more service interruptions, all of which negatively affect customer confidence.

Ultimately, production is where a solution spends the majority of its life and where it generates value for the customer. A project that reaches go-live without being production ready transfers technical debt directly into the operational phase, where it becomes more expensive and more disruptive to resolve.

The Solution

SRE should become an active stakeholder throughout the delivery lifecycle rather than a gatekeeper at the end.

The objective is not for SRE to own delivery activities, but to provide operational guidance early enough that production readiness becomes part of the project's definition of done.

A more effective delivery model looks like this:

Proposal
    │
Discovery
    │
SRE Engagement
    │
Solution Design
    │
Delivery
    │
Production Readiness Reviews
    │
Go Live
    │
SRE Onboarding
    │
Delivery Team Rolls Off
    │
SRE Operates the Service

Engage SRE During Discovery and Design

SRE should participate during discovery and solution design to review operational requirements before implementation begins.

This allows operational concerns such as monitoring, resilience, supportability and disaster recovery to influence architecture while changes are still inexpensive.

Define Production Readiness Up Front

Production readiness requirements should be agreed at the start of the project and tracked alongside functional requirements.

Areas that should have defined acceptance criteria include:

  • Observability
  • Monitoring and alerting
  • Logging
  • Backup and disaster recovery
  • Security
  • Performance and capacity
  • Operational documentation
  • Runbooks
  • Deployment processes
  • Support model
  • Operational ownership

These requirements should form part of the project's acceptance criteria rather than being treated as optional enhancements.

Perform Regular Production Readiness Reviews

Rather than conducting a single assessment immediately before handover, SRE should perform periodic production readiness reviews throughout delivery.

This enables issues to be identified while there is still sufficient time and budget to address them.

Treat Production as a Primary Deliverable

Delivery projects naturally focus on building features. However, the long-term success of a platform depends on how well it performs in production.

Production readiness should therefore be considered a first-class deliverable alongside functional capability.

A solution cannot be considered complete simply because it works. It should also be observable, supportable, secure, resilient and maintainable.

Make Production Readiness a Shared Responsibility

Operational excellence should not be viewed solely as an SRE concern.

Delivery teams, architects, developers, testers, infrastructure engineers and SRE all contribute to producing a platform that is ready for long-term operation. Shared ownership throughout delivery results in fewer late surprises and a smoother transition into managed service.

Benefits

Earlier SRE engagement provides measurable benefits for both delivery teams and operational teams, including:

  • Fewer production readiness issues identified before go-live.
  • Reduced operational risk.
  • Faster and more predictable onboarding into SRE.
  • Lower operational overhead.
  • Improved service reliability and availability.
  • Faster incident detection and resolution.
  • Reduced long-term operational cost.
  • Greater customer confidence in the live service.

By shifting production readiness activities earlier in the delivery lifecycle, organisations avoid transferring operational debt into managed services and create platforms that are significantly easier to support throughout their operational lifetime.