I was in a hotel room in Vienna, laptop open, headphones in, watching a recording of myself performing at a corporate event from two weeks earlier. I had my script open in a document next to the video. The plan was simple: follow along with the script as I watched myself perform, checking that the words matched.
They did not match.
Not even close.
My script said: “I want you to think of someone important to you. Someone whose opinion matters. Hold that person in your mind.”
What I actually said, according to the recording, was: “Okay so what I want you to do is, um, I want you to think of someone. Someone who’s important to you, right? Like someone whose, whose opinion really matters to you. So just, you know, hold that person in your mind for me.”
Same idea. Twice the words. Littered with filler. Peppered with verbal tics I had no idea I was producing. The “okay so” at the beginning. The “um.” The “right?” thrown in as a nervous question. The repeated “whose, whose” stutter. The “you know” that added nothing. The “for me” tacked onto the end like an apologetic afterthought.
I was not delivering my script. I was delivering a loosely related improvisation based on the general theme of my script. And until I watched that recording with the written script open beside it, I had no idea.
This is what Ken Weber describes in Maximum Entertainment as the “reverse engineering” technique: you transcribe what you actually say during a performance — from video, word for word — and then compare it to what you think you say. The gap, he warns, will be larger than you expect.
He was right. The gap was enormous.
The Script You Think You Deliver
Here is the thing about having a memorized script. You know the words. You have rehearsed them. When someone asks you what you say during a particular routine, you can recite the script from memory. You believe, sincerely and completely, that those memorized words are the words leaving your mouth during performances.
But there is a difference between knowing the words and saying the words. Under the pressure of a live performance — with an audience watching, with the stakes of the moment pressing down on you, with your brain simultaneously managing the technical aspects of the effect and the social dynamics of the interaction — your delivery drifts. Not dramatically. Not in ways that anyone in the audience would notice. But it drifts.
You add filler words. You repeat yourself. You insert qualifiers that are not in the script. You soften statements that were written as strong declaratives. You rush through the lines you are less confident about and linger on the ones that feel safe. You develop verbal habits — patterns of speech that emerge under pressure and repeat themselves across every performance without your conscious awareness.
These are not mistakes in the traditional sense. Nobody in the audience thinks you stumbled. The performance feels coherent. The words make sense. But they are not the words you chose when you sat down and carefully crafted your script. They are the words your subconscious produced in the heat of the moment, using your script as a rough blueprint rather than an exact specification.
You have a subconscious script. And it is overriding your conscious one.
The Transcription Exercise
After that first humbling comparison in Vienna, I decided to do the exercise properly. Not just for one routine, but for my entire set.
I set up my phone to record a performance at a private event in Graz. Then I came back to my hotel, opened a blank document, and transcribed myself word for word. Every syllable. Every “um.” Every false start. Every repeated phrase. Every “so” and “just” and “okay” and “right?” that I produced unconsciously.
It took three hours to transcribe a thirty-minute set. Three hours of hearing myself say things I had no memory of saying. Three hours of watching my carefully crafted sentences dissolve into rambling approximations of themselves.
The patterns were immediately obvious.
First, the filler words. I used “so” as a crutch at the beginning of almost every sentence when speaking to a volunteer. “So what I’d like you to do.” “So here’s what’s going to happen.” “So if you could just.” The word “so” was not in my script. Not once. But in performance, it preceded roughly forty percent of my sentences like a nervous tic.
Second, the qualifiers. My script was written in declarative sentences. “Think of a number.” “Look at the card.” “Remember this moment.” Clean, direct, confident. But in performance, I softened almost every directive with qualifiers. “I’d like you to think of a number.” “If you could look at the card.” “I want you to remember this moment.” The shift from declarative to conditional was subtle but pervasive. Instead of telling people what to do, I was asking them if they would mind perhaps considering the possibility of doing it.
Third, the padding. Phrases that contributed nothing but occupied time. “What I’m going to do now is.” “What’s interesting about this is that.” “The reason I mention that is because.” These are throat-clearing phrases — the verbal equivalent of shuffling papers before starting a meeting. They signal that content is coming without actually delivering any content.
Fourth, the nervous questions. “Does that make sense?” “You got that?” “Okay?” “Right?” Scattered throughout my performance like anxious checkpoints, each one a small confession that I was not fully confident the audience was following me. Each one an interruption to the flow of the performance. Each one an unnecessary moment where I was asking the audience for reassurance rather than leading them with certainty.
What the Subconscious Script Reveals
The transcription does not just show you what you say. It shows you what you feel.
When I studied my transcript, I could see my confidence levels mapped across the performance like a topographic chart. The sections where I was comfortable — well-rehearsed material, pieces I had performed dozens of times — were clean. Relatively few filler words, declarative sentences, minimal padding.
The sections where I was less confident — newer material, transitions, moments where I was managing a volunteer interaction I was not sure would go smoothly — were dense with verbal debris. “So,” “um,” “right?”, “just,” “actually,” “kind of.” The filler words clustered around the moments of uncertainty like barnacles on a ship’s hull.
This was useful information. Not because I could magically eliminate the uncertainty — some of that comes with experience — but because it showed me exactly where my script needed more work. The sections with the most verbal debris were the sections I had under-rehearsed. They were the sections where my conscious script was not deeply enough embedded to override my subconscious defaults.
The fix was not to try harder in the moment. The fix was to go back to rehearsal and drill those sections until the conscious script was strong enough to hold under pressure. Until the words I chose deliberately were more automatic than the words my anxiety produced.
The Verbal Tics You Cannot Hear
Here is what I find most unsettling about this exercise: I could not hear my verbal tics in real time. During the performance, I had no awareness of the “so” that preceded every sentence. I had no awareness of the “right?” that punctuated every instruction. I had no awareness of the qualifiers that softened my directives.
This is because your brain filters out your own speech patterns. You hear the content of what you are saying, not the texture. You hear the meaning, not the filler. It is similar to how you stop hearing the hum of an air conditioner after a few minutes — the brain categorizes it as irrelevant background noise and removes it from conscious awareness.
Your filler words are your air conditioner. You stop hearing them. But the audience does not. The audience hears every “um,” every “so,” every “right?” And while individual filler words are imperceptible, their cumulative effect is real. They make you sound less confident, less prepared, less authoritative. They are tiny cracks in the facade of competence that, collectively, undermine the audience’s faith that they are in capable hands.
The only way to hear your verbal tics is to put them in writing. When you transcribe your own speech, every filler word is right there on the page, undeniable and impossible to filter out. You cannot not-see the word “um” when it is written in black text. You cannot ignore the “right?” when it appears seventeen times in a transcript.
The Comparison Process
Once I had a full transcription, I printed it alongside my intended script and compared them line by line. I used three colors. Green for lines that matched the script closely. Yellow for lines that were in the right territory but had drifted in wording or structure. Red for lines that had diverged significantly or that contained material not in the script at all.
The result looked like a traffic light having a bad day. Mostly yellow and red, with islands of green where I had been most confident.
But here is the productive part. For every yellow and red line, I asked a simple question: is the version I actually said better or worse than the version I scripted?
Sometimes the improvised version was better. Sometimes, in the heat of performance, I found a phrasing that was more natural, more rhythmic, more connected to the moment than what I had written at my desk. In those cases, I updated the script to reflect the improvement. The performance had accidentally edited the writing, and the edit was an upgrade.
More often, the improvised version was worse. Longer, weaker, less precise, cluttered with filler. In those cases, I went back to the original script and drilled it harder. I rehearsed those specific lines with extra repetition, extra intention, extra focus on hitting the exact words I had chosen.
Over time, through multiple cycles of perform-transcribe-compare-adjust, my actual delivery got closer and closer to my intended delivery. The gap between the conscious script and the subconscious script narrowed. Not to zero — it will probably never be zero, because live performance always introduces some drift — but to a margin that I could live with.
A Recurring Practice
I do this exercise every few months now. Not because I think I have backslid, but because new tics develop over time. The “so” crutch was something I mostly eliminated, but six months later I noticed I had started using “look” as a transitional word. “Look, here’s the thing.” “Look, I want to show you something.” I had traded one tic for another without noticing.
The transcription exercise is not a one-time fix. It is a recurring diagnostic. Like getting your car checked — you do not do it once and assume everything is fine forever. You do it regularly, because things drift, things wear down, and new problems develop that you will not notice until you look under the hood.
The process is not fun. Watching yourself on video is uncomfortable. Transcribing your own speech is tedious. Comparing the transcript to the script is humbling. But it is the most direct path I have found from performing what I think I am performing to actually performing what I intend to perform.
The subconscious script is not your enemy. It is a signal. It shows you where your preparation is strong and where it is weak. It shows you where your confidence holds and where it cracks. It shows you, with merciless clarity, the difference between the performer you imagine yourself to be and the performer the audience actually sees.
That gap is where the work is. And the first step to closing it is to see it.