Our favourite eggs to break · 02
Change, configuration & resilience.
Every outage review asks the same question: what changed? If the honest answer is “we're not completely sure”, the problem isn't the people. It's a change and configuration process that has drifted away from how your estate really works.
Sound familiar?
- Incidents that trace back to changes nobody can fully reconstruct: what changed, when, and who approved it.
- A change advisory board that reviews paperwork rather than risk, and approves almost everything anyway.
- Emergency and “standard” changes quietly becoming the normal route for anything urgent.
- Configuration drift between environments that only shows up when something fails.
- Auditors and regulators asking for evidence that takes weeks to assemble by hand.
- Operational resilience obligations (DORA, NIS2, FCA) landing on teams that were already stretched.
Why it happens
Change processes are usually designed for the estate you had ten years ago: fewer systems, slower releases, clearer ownership. As delivery speeds up and the estate fragments across cloud, SaaS and on-premise, the process keeps its shape while the reality underneath it changes.
The result is a process people work around rather than with. Approvals become a formality, the real risk conversations happen elsewhere, and the evidence trail regulators now expect simply isn't being created. It's “how we've always done change”, so nobody wants to be the one to question it.
What it's really costing you
- Avoidable incidents and the customer and reputational damage that comes with them.
- Slow delivery, because the process adds friction without adding safety.
- Expensive, last-minute scrambles to produce audit and regulatory evidence.
- Exposure to findings and enforcement as resilience regulation tightens.
How we break this egg
Look at what actually happens
Trace real changes and incidents end to end, so we see the process people follow rather than the one in the policy document.
Right-size the risk model
Separate genuinely risky change from routine work, so effort goes where the risk is and low-risk change flows faster.
Connect change to configuration
Make sure every change can be tied to what it touched and what depends on it, so impact and drift are visible before they bite.
Build evidence in, not on
Design the process so the audit trail is produced as a by-product of doing the work, ready for auditors and regulators when they ask.