Delete the content step you cannot defend to anyone outside your team, then ship the next ten posts without it and see who complains. Almost nobody will. That is the whole tactic, and it costs nothing to run this week.
Source: Alex Hormozi's 2026-08-12 video "My Take on Elon Musk's First Principles Framework | Creating Content & Personal Branding". The color-grading anecdote and the three questions are his. James Smith's iPhone test on the same theme is corroboration, not a second tactic. Framing this as a content-ops process audit rather than an editing debate is ours.
Every content operation accumulates undefended steps
Every content team accumulates steps that nobody chose. Somebody said years ago that reels needed brand-color captions, somebody else said the founder needed to review every hook, and now the step is in the pipeline because it has always been in the pipeline. You do not have a document that says why. You have a habit, and everyone assumes someone else knows the reason.
Hormozi's team hit one of these. He asked what was eating the most hours in the operation, someone said color grading every video, he asked who decided that was required, and the room went quiet. Nobody had an answer, and the step had been standard for long enough that questioning it felt slightly rude. That is usually the tell. The steps you cannot audit out loud are the steps most likely to be dead weight.
Elon's rule, running on a content pipeline
The move Hormozi ran is not new. He is applying a rule Elon Musk uses in manufacturing: delete parts of the process until you have to add something back, then only add back what you actually needed. If nothing breaks, the step was never load-bearing. If something breaks, you learn exactly what the step was doing, which is often narrower than the story you told yourself about it.
Applied to a content pipeline it looks like this. Take one required step in your process (the extra review round, the caption rewrite, the "we always color grade" rule, the founder-approves-every-hook gate), skip it on the next batch, and see whether anything measurably drops on the metrics you already track: views, saves, replies, whatever your dashboard shows. Treat the step as innocent until the data convicts it, not the other way around.
The three questions Hormozi ran on "we color grade every video"
Alex ran three questions on anything the team marked as required.
- What does that mean?
- How do you know that?
- So what?
Color grading did not survive those three questions on his account. When they went back through past performance, some of his best videos were not graded and some of the videos that flopped were. The step had not been earning its place, it had been drafting off a general belief that polish equals performance. So they cut it, watched, and it turned out nobody outside the team could tell the difference.
There is a cost lens under this that is easy to miss. Hormozi framed grading as: if it eats 40% of the hours the team spends per video, it has to add 40% of the quality to pay for itself. He was explicit that the numbers were a thought experiment, not measured data, so treat them as a way to think about the trade rather than a benchmark. But the shape of the question holds up in most content operations we have looked at. Steps you cannot defend on cost you also cannot defend on quality, because the person asked to defend them has never checked.
A second data point from a different creator
James Smith ran a version of the same test recently and came to the same conclusion on his grid. Polished, produced videos were sitting around 40 to 70 thousand views. One iPhone cut he shot without the usual setup did 640 thousand. Those are on-screen Instagram counts rather than audited analytics, so read them as his lived experience rather than a benchmark, but the direction is consistent. Two creators at very different scales, cutting different steps, landing on the same finding: the polish was optional and nobody had proved otherwise.
The reason both examples matter is that the tactic is a process audit, not a production-budget argument. The specific step being questioned changes in every operation. Yours might be a review round, a B-roll rule, a caption rewrite, a founder-approves-the-thumbnail gate, a "we always add captions in the brand font" rule. What matters is that you can name it, that nobody in the room can produce a piece of evidence for it, and that you are willing to run without it for long enough to see whether anything actually breaks.
How to run this on your operation this week
Pick one Friday afternoon. It takes about an hour.
- Write down every step between "idea" and "posted." Not the summary, the full list. Circle anything you cannot immediately defend with a specific result.
- Skip one of the circled steps for the next ten posts. Label those ten so you can find them again later, a tag, a saved view, whatever your tool supports.
- Compare that batch to the ten posts before it on the metrics you already care about. If nothing measurably dropped, leave the step dead or make it optional. If something did drop, you have learned what the step was actually doing, which is usually narrower than what you assumed.
The reason this works is that most content pipelines have never been audited. They were assembled a step at a time, usually by people who are no longer on the team, and nobody has the political air cover to question the parts that feel professional. Running the audit on ten posts is small enough to be safe and big enough to be real.
The steps you inherited are not neutral
Content teams talk about process the way engineering teams used to talk about legacy code, with a mix of respect and resignation. It works, nobody wants to touch it, and everyone assumes it is load-bearing until they try to remove it and find out it was scaffolding somebody put up during a specific problem three years ago. Every quarter your process should shed a few of those, because the audit is the only way to find out which parts of the operation are actually earning their place and which parts are just old.
