Bilbo’s sword Sting has one magical property: it glows blue when orcs are near. Not when they attack — when they approach. The whole value is in the gap between the glow and the danger. Bilbo gets his warning while there is still time to act.
Most transformation programmes have dashboards full of indicators that report problems after they arrive. Red status reports are not early warnings. They are obituaries. By the time a milestone officially slips, the causes are weeks old.
Rehearsing the worst day before it comes
Early in the 2000s I was responsible for IT security and continuity at a large financial-services group — before anyone used the word CISO. We survived a real worm attack, and the organisation learned plenty from it. But the work I still think about is what we did before anything happened: we rehearsed. Full crisis simulations, up to a scenario in which a plane crashes on headquarters. Managers convened offsite, stood up a provisional office, and worked the real questions under time pressure — who is dead, who may talk to the press, who must not. A colleague played the press and needled everyone relentlessly. It was uncomfortable by design.
Years later, in a different organisation entirely — a large IT services provider — I built a SWAT function for rescuing troubled projects. People ask which crisis triggered it. The honest answer: none. I built it because I had learned that the muscle has to exist before the fire.
And the pattern proved itself a third time, years later again and in yet another setting. When a real intrusion forced the full shutdown and rebuild of a client’s data centre — as political and high-stakes as incidents get — the response was calm and coordinated, decisive and focused, run by a team whose trust had been built long before the alarm sounded. Three different companies, across three decades of my career, and the same principle every time: every crisis I have seen managed well was won in advance. Sting is no use if you only go looking for a sword once the orcs arrive.
Sting metrics
Readiness is one half. The other half is knowing when to reach for it. Over the years I have come to watch a set of signals I think of as Sting metrics. They come not from the project plan, but from how the team is actually behaving:
- A sudden drop in team velocity, before anyone has reported a blocker.
- Rising email volume between two stakeholder groups. People CC-ing for self-protection is one of the most reliable conflict signals there is.
- Status reports arriving later, or growing vaguer.
- Meetings that used to produce decisions starting to produce follow-up meetings instead.
None of these appear on a conventional RAG dashboard. All of them are measurable. And, like Sting, they glow blue while there is still time to act — usually weeks before the problem shows up in red. The hard part is never collecting them. It is taking them seriously while the official dashboard is still comfortably green.
What this means for your transformation — and your AI adoption
The principle is simple: instrument behaviour, not just deliverables. Deliverables lag. Behaviour leads.
This matters double for AI right now. Boards adopting AI are building governance dashboards full of lagging indicators: incidents, audit findings, compliance attestations. All necessary, all obituaries. The governance that actually protects you watches the leading signals: which teams have quietly started using unapproved tools, where staff are overriding model outputs (or worse, never questioning them), which decisions have drifted from human to machine without anyone deciding they should. That is Sting work, and it is governance craft, not technology.
The best leaders I have worked with share one trait: they sense danger before it strikes, and they act on the glow rather than waiting for proof. Proof always arrives. It just arrives too late to be useful.
“Never laugh at live dragons.”
— The Hobbit, Chapter 12
Part 6 of ten in There and Back Again — Transformation Lessons from an Unexpected Journey. Previous: Barrel Riding. Next: The Burglar’s Role — on why the outsider sees what insiders cannot.
Leave a Reply