I overran a keynote in Salzburg by eleven minutes.
Not through carelessness. The effects ran at the expected length. The scripted sections delivered at roughly rehearsed time. What I hadn’t accounted for was the rest of it: the reaction that ran longer than I expected, the volunteer interaction that took an extra three minutes, the moment the room needed longer to settle than my mental program had budgeted.
At the forty-minute mark, I was looking at about twenty minutes of material I needed to deliver in nine. The organizer was standing at the side of the room with the specific expression that means “you’ve already cost me someone’s lunch break.” I started cutting. On the fly, mid-performance, deciding what to drop and what to compress while simultaneously performing the thing I hadn’t dropped yet.
I got out in forty-nine minutes. The show ended. The organizer thanked me in a way that was technically a compliment.
I drove home and wrote down the rule I had failed to follow, and have followed ever since: design for ninety percent of the time you have.
Where the Missing Time Goes
The first thing you need to understand about show timing is that rehearsed time and performed time are different numbers, and the gap between them almost always goes the same direction.
In rehearsal: you control the pace. There’s no audience. The beats are clean, the transitions are immediate, the scripted sections run at their text length. Everything is optimized for efficiency because efficiency is possible when nothing is actually happening.
In performance: the audience exists, and the audience is not optimized for efficiency. They laugh when you didn’t expect them to laugh, and the laugh runs as long as it runs. They ask questions you weren’t planning for. The volunteer who comes up needs thirty seconds to collect themselves before they can participate. The effect that always took forty-five seconds takes seventy because the reaction ran through and then ran again. The organizer introduces you for two minutes instead of thirty seconds.
None of this is failure. All of it is performance — the actual living version of the thing you rehearsed the static version of. But it takes time. And if your show is designed to exactly fit the allotted slot in rehearsal, it will almost certainly exceed that slot in performance.
The Ninety Percent Rule
The arithmetic is simple: multiply your total time by 0.9, and design a show that can be delivered comfortably in that adjusted duration.
For a forty-five-minute slot, design for forty. For sixty minutes, design for fifty-four. For thirty minutes, design for twenty-seven.
This creates a buffer — ten percent of the total slot — that absorbs the natural expansion of live performance. If the effects all run at their rehearsed length, if no unexpected moments occur, if the audience does nothing surprising: you finish a few minutes early, which is the correct outcome. Audiences and organizers will not complain about a show that ended slightly before it was supposed to. They will complain, in the particular way that involves scheduling fallout and passive-aggressive emails, about one that ran over.
Ken Weber writes about the director’s perspective: the professional who can control time within their performance is demonstrating a quality of craft that amateurs don’t have. Time management is not a side issue to the craft of performance. It is a central expression of it.
What Rushing Costs
The failure mode of overrunning isn’t just organizational — it changes the show.
When I realized in Salzburg that I was eleven minutes over budget with nine to go, I started making decisions I would not have made otherwise. Cut this effect. Compress that section. Skip the moment I normally take to let the previous effect land before moving forward.
These decisions show. Not necessarily as “the performer is rushing” in an obvious, clumsy way. But as a subtle shift in tone — a slight tightening of the pace, a loss of the spaciousness that makes a show feel generous rather than pressured. The breathing room disappears. The show starts to feel managed rather than inhabited.
The audience can sense this shift even if they can’t articulate it. The experience of being in the presence of a performer who is slightly behind is different from being in the presence of a performer who is exactly where they expected to be. The second one is more interesting to watch. They seem more authoritative. More in control.
Because they are. The control is structural: their design gives them room, and room feels like authority.
Building the Buffer Correctly
The practical question is how to build the buffer without feeling like you’re leaving material out.
There are two approaches I’ve used, and they work in different contexts.
The first is elimination. Choose which effects or sections are lowest priority, and remove them from the designed show. The show is shorter by design, not by in-the-moment cutting. You know what you’re performing, and the eliminated material isn’t in your head competing for space.
The second is expansion points. Keep the full show but identify specific sections where you can expand if time allows or compress if it doesn’t. These are typically the moments of audience interaction or the beats where a reaction could run long — places where the duration is variable by nature. In rehearsal, you mark these as “can shorten to X minutes, could extend to Y minutes.” You make the decision in real time based on where you are in the program.
I use both, depending on the show. For tight keynote slots — thirty minutes or under — I eliminate. For longer shows with more flexibility, I design expansion points.
The Organizer Relationship
One underrated benefit of the ninety percent rule is what it does for your relationship with event organizers.
Organizers live and die by the schedule. The speaker who runs over is a problem that cascades: the panel that was supposed to start at 11:15 now starts at 11:26, which means lunch runs late, which means the afternoon session starts late, which means the day’s final session gets compressed. One overrun creates downstream damage the organizer has to manage for hours.
The performer who consistently ends on time or slightly early builds a specific kind of professional reputation. Not flashy. But solid. The kind of performer that organizers invite back because they don’t create problems. In competitive speaker and performer markets, reliability is an undervalued differentiator.
I’ve been invited back to events not because my show was the most impressive in the slot, but because I was the person who said “forty-five minutes” and delivered forty-three and a half. Organizers notice.
The Eleven Minutes
That overrun in Salzburg cost me something beyond the organizer’s goodwill. It cost me the ending of the show.
The final piece I had planned — the one designed to land the emotional note the whole talk had been building toward — got cut in the scramble. What I delivered instead was a compressed version of something that needed full space to work. The audience was polite. The real ending was somewhere on the cutting room floor.
Design for ninety percent. Protect the ending.
The material you couldn’t use because you gave yourself room isn’t wasted. It’s waiting for the next show, in its correct position, with the time to do what it was designed to do.