From QA to Release Management: What 2025 Taught Me About Quality

The past few years didn’t feel like a career change. They felt like a lens change.

I didn’t wake up one morning and decide I’m “done with QA.” I’m still made of QA - skepticism, curiosity, the reflex to trace outcomes back to assumptions. What shifted was the shape of the work: from “I help teams build quality into what we ship” to “I design how we ship without breaking the organisation.”

Being in QA trained me to think in systems: intent, risk, feedback loops, and the human incentives that quietly decide outcomes. Release management trained me to turn that thinking into decisions—and to feel, in real time, the hidden cost of getting them wrong.

If my earlier identity was quality advocate, my 2025 identity became path designer: building routes to production that are fast when they can be, deliberate when they must be, and honest about the trade-offs either way.

Lesson 1: QA gave me the quality lens. Release management gave me the steering wheel.

In QA, you can influence outcomes through evidence: you surface risk, you challenge assumptions, you add signal. You make the invisible visible.

In release management, you’re often responsible for turning that visibility into coherent movement:

• aligning timelines across teams,

• making dependencies explicit,

• negotiating scope without losing control of risk,

• ensuring operational readiness isn’t an afterthought,

• and keeping the organisation honest about what it’s actually shipping.

This year taught me a subtle distinction:

QA is excellent at generating signal.
Release management is about reducing decision latency—getting the right people to make the right call at the right time, with the right constraints.

That’s not “bigger” work. It’s just work at a different leverage point.

Lesson 2: Quality work became more about decision design than decision drama.

I didn’t stop caring about defects. I stopped treating defects as the primary unit of truth.

At scale, “quality” is not a list of bugs. It’s a pattern of decisions:

• what we accept,

• what we postpone,

• what we mitigate,

• what we monitor,

• and what we refuse to normalise.

A release isn’t a single moment. It’s a chain of commitments.

So I started obsessing less over whether something was perfect and more over whether we had:

• a clear risk narrative,

• bounded blast radius,

• rollback readiness,

• a monitoring plan,

• and a realistic story about what could go wrong and how we’d know.

That’s not lowering the bar. That’s moving the bar to where it actually lives: in the socio-technical system, not in the spreadsheet.

Lesson 3: “Not yet” is a governance tool. “No” is a posture.

I’ve always disliked the gatekeeper identity—not because boundaries don’t matter, but because “no” can become a performance: clean, satisfying, and strangely unhelpful.

Release work forced me to develop a more precise vocabulary:

• “Not yet, because…” (grounded in risk, not preference)

• “Yes, if…” (with explicit mitigations and owners)

• “Yes, but…” (with transparent trade-offs)

• “No, unless…” (with honest accountability)

This is where “quality is shared responsibility” stops being a slogan and becomes operational reality. Shared responsibility needs shared language. Otherwise, it’s just shared confusion.

Lesson 4: ITIL isn’t the enemy. Untranslated governance is.

I’ve seen two kinds of dysfunction:

1. Governance-as-theatre: approvals as superstition, templates as shields, compliance as a substitute for thinking.
2. Agility-as-vibes: speed worship, unclear ownership, and “we’ll deal with it later” as a delivery strategy.

Neither is sustainable.

ITIL concepts—change enablement, risk acceptance, incident/problem learning—can coexist beautifully with continuous delivery when they’re treated as outcomes, not rituals.

My job became translation:

• translating delivery reality into business risk,

• translating governance intent into team behaviour,

• translating “we need control” into “we need reliable feedback loops and explicit guardrails.”

When ITIL is done well, it doesn’t slow teams down. It stops teams from paying interest on invisible debt.

Lesson 5: CAB is where quality becomes social—sometimes uncomfortably so.

CAB conversations are rarely about “is it good?” They’re about:

• who owns the risk,

• whose urgency gets prioritised,

• whose uncertainty is tolerated,

• and how power moves through the organisation.

I learned quickly that the hardest part isn’t the mechanics. It’s the human layer:

• confidence substituting for evidence,

• urgency flattening nuance,

• accountability drifting sideways.

The growth edge for me was learning to operate there without becoming cynical:

• insist on clarity without humiliating people,

• keep risk language concrete,

• avoid turning governance into theatre,

• and protect teams from chaotic, last-minute whiplash.

CAB is not just a meeting. It’s a mirror.

Lesson 6: Visibility isn’t vanity. It’s part of resilience.

There’s a romantic impulse in QA to be the quiet professional: do the right work, stay out of the spotlight.

But invisibility has a cost. If nobody can see the work, they assume it’s automatic. If it looks automatic, it becomes endlessly expandable. And that’s how you quietly become a single point of failure.

So I learned to make the work visible in a way that’s useful, not performative:

• crisp comms,

• explicit decisions,

• documented assumptions,

• predictable cadence,

• clear ownership.

This wasn’t personal branding. It was organisational memory—and self-preservation.

Lesson 7: The emotional load of owning releases is real, and it deserves respect.

Release work has a strange asymmetry: you often prevent disasters you can’t prove would have happened.

You carry uncertainty so others can keep building. You absorb tension so teams can stay focused. You get measured on outcomes that are partly outside your control, and you’re expected to make it all look effortless.

In 2025, I stopped pretending that’s “just part of the job” in the hand-wavy sense.

I started treating emotional load as an operational factor:

• boundaries,

• sustainable cadence,

• fewer heroic interventions,

• and more engineered guardrails.

The hero narrative is expensive. It bills your future self.

Why I started Null Expected (and why it mattered more than I expected)

Null Expected began as a thought hub because I wanted a space where quality isn’t reduced to either:

• defect counts, or

• motivational posters about “quality culture.”

Writing publicly did something useful to my career: it forced precision.

You can have instincts for years. Writing makes you articulate the underlying model:

• what you believe,

• why you believe it,

• and what you’d do differently next time.

It also made me reclaim a narrative I care about: I’m not “ex-QA.” I’m still doing quality work—just through a wider, more organisational lens.

What I’m leaning into in 2026

I don’t want 2026 to be a year of vague ambition. I want it to be deliberate.

• More thought leadership with teeth: fewer platitudes, more real constraints and trade-offs.

• Release governance as a designed system: scalable, humane, outcome-driven.

• Stronger risk language across teams: fewer “it should be fine,” more explicit assumptions and mitigations.

• Conference talks + public writing: not for noise—because it sharpens the craft.

• Sustainability over heroics: build guardrails; stop relying on adrenaline.

If 2025 taught me anything, it’s this:

Quality isn’t a department and it isn’t a personality trait.
Quality is the discipline of making risk visible, decisions explicit, and outcomes sustainable—over time, with humans in the loop.

Release management didn’t change what I value.
It changed where I apply it—and how much leverage that gives me.

Disclaimer: The perspectives expressed herein are personal interpretations intended to foster professional dialogue; they do not represent any official stance of current or former employers.