Every event team has a story about the last-minute heroics — the team that solved the broken broadcast feed at 14:30 the day of, the producer who rebuilt the run-of-show on the morning of the keynote, the AV engineer who replaced a failed rig in 40 minutes. These stories get told. They are also the wrong stories.
The honest version: most events that fail were going to fail by T-7. The heroics on the day are damage limitation, not save. By the time you're in the venue on day-zero, the project has either accumulated enough operational tension to absorb surprises, or it hasn't. The window to fix it is the pre-event week — and the week before that.
The four pre-event slip signals
We've run enough events now to recognise the early warnings. Four signals show up in the T-21 to T-7 window. Any one of them on its own is manageable. Two of them together is a yellow flag. Three is a project that's going to slip publicly.
Signal 1: The run-of-show is still in flux at T-10
By T-10, the run-of-show should be locked. Speakers know what they're doing, when, with what cue. AV knows the order of cues. The producer knows the failover plan for each beat. If at T-10 the agenda is still being moved around — particularly if leadership keeps adding things — the on-the-day will be characterised by frantic last-minute changes that the venue team can't absorb.
The fix is to push back at T-21 with a frozen run-of-show and require a written change-control for every subsequent edit. This is unpopular. It is also the single most reliable predictor of an event landing well.
Signal 2: Speaker prep is "scheduled for closer to the date"
If by T-7 the leadership speakers haven't done a tech rehearsal, you have a serious problem. They will arrive on the day with no muscle memory for the stage, no comfort with the cue, no feel for the broadcast. The first take of the keynote will look like a first take.
The fix is mandatory tech rehearsals at T-7 — even short ones, 20 minutes per speaker. Not because the leadership team can't perform, but because the room and the broadcast need to know the speakers' patterns.
By T-7, you can predict whether an event will land well with about 80% accuracy. The day-of heroics are damage limitation, not save.
Signal 3: Vendor list isn't single-threaded
If at T-14 you're still chasing seven different vendors for confirmations — caterer, AV, badge printer, branded merch, photography, security, transport — and none of them have a single accountable lead, you're managing a coordination problem that will absorb the producer's full attention on the day. The producer should be running the show, not managing logistics.
The fix is to consolidate all vendor lines under one ops lead by T-21 — even if that means paying a small coordination fee. The cost of an ops lead is significantly less than the cost of a producer dropping a cue because they were on the phone to the catering team.
Signal 4: Engagement systems aren't in dry-run yet
If at T-5 the audience-response system, the registration flow, the sponsor dashboards, the recording pipeline haven't all been dry-run end-to-end, expect on-the-day surprises. Every system needs to be tested under representative load before the day. The dry-run isn't a scheduling formality. It's a load test.
The pre-event audit
We run a structured pre-event audit on every Zlicc engagement. It looks like this:
- T-30: Run-of-show v1 locked. Speaker list locked. Vendor consolidation done.
- T-21: First full dry-run with leadership speakers. Sponsor onboarding complete. Microsite live.
- T-14: Run-of-show v2 frozen. Change-control active. Tech rehearsal schedule confirmed.
- T-7: Final dry-run at venue conditions. All engagement systems live. Failover paths tested.
- T-2: Stage call. Final speaker walk-through. Run-of-show printed.
- Day 0: Show.
Three engagements out of every five we've run got tightened by this audit at T-7. One out of five was already on track. One out of five surfaced an issue that would have caused a public failure. None of them produced a day-of disaster — because the disasters were caught when they were still cheap to fix.
The wrong story
The on-the-day heroics make for better case studies than the boring T-21 sign-off. But the audit is the work. The heroics are the consequence of the audit not happening. Every team should be measured on how few heroics it produced, not how many.