Exercise 1: Why Data Loss Needs Extra Communication Care — Possible Solution ==================================================================== WHAT MAKES A TYPICAL SERVICE OUTAGE DIFFERENT ------------------------------ Per this chapter, "'the service will be back in an hour' is bounded and recoverable - uncomfortable, but temporary." A service outage, however disruptive, carries an implicit promise: whatever's wrong now will eventually be fixed, and normal service will resume. Even a bad update delivered during an outage can be corrected later once the situation improves. WHAT MAKES DATA LOSS STRUCTURALLY DIFFERENT ------------------------------ Per this chapter, "'this specific data cannot be recovered' carries a permanence other incident types don't share." Once specific data is genuinely gone beyond what any backup can restore, no later update, no matter how well-handled, can undo that fact. The finality isn't about how the incident is communicated - it's a property of the underlying situation itself. WHY THIS MEANS THE STAKES OF AN EARLY MISTAKE ARE HIGHER ------------------------------ Per this chapter, "there's no later update that undoes a wrong answer given too early." In a typical outage, an overly optimistic early estimate can be corrected with an apology and a revised timeline, and the underlying problem still gets fixed eventually. In a data-loss incident, an overly optimistic early statement about recovery odds sets someone up for a second, avoidable disappointment once the truth becomes clear - the damage from a badly handled message compounds on top of a loss that was already permanent. WHY incident1's OWN PRINCIPLES STILL APPLY, JUST WITH MORE CARE ------------------------------ This chapter doesn't argue that data loss needs an entirely different communication approach - staged updates, honest uncertainty, and audience-appropriate framing (all `incident1`'s own territory) still apply directly. What changes is the margin for error: because a wrong early statement about data loss can't be quietly corrected the way a wrong outage-duration estimate can, the same principles need to be applied more carefully, not replaced. WHY THIS WORKS AS AN ANSWER ------------------------------ It contrasts the recoverable, bounded nature of a typical outage with the permanence of data loss, explains why that permanence raises the stakes of an early mistake specifically, and explains why this means applying incident1's existing principles more carefully rather than needing an entirely different approach.