The highest-earning affiliate video we have ever analysed made one point.
Rp 119 million in sales, 24 seconds long, and the entire script rested on a single idea: you plug it in. That was the message. Not the two-mode switch, not the 45-day refill life, not the absence of smoke or ash. Those appeared, but late, and as supporting detail.
The videos that sold nothing said more. They named the efficacy claim, the duration, the cost saving, the smoke-free benefit, and the mode options, all inside thirty seconds. Every claim was accurate. Nobody remembered any of them.
This is the inversion most content operations get wrong. They say a hundred things once. The work that compounds says one thing a hundred times.
What the Failure Actually Looks Like
We audited 5,871 affiliate videos across nine FMCG product lines in Indonesia and ran frame-by-frame teardowns on the extremes. The pattern held in every category.
Top performers locked one differentiator in seconds two through six and spent the rest of the runtime proving it. Failures spread four or five claims across the same window and proved none of them.
Duration tells the same story from another angle. Winners averaged 31 seconds. Zero-revenue videos averaged 37. The extra six seconds were not adding evidence. They were adding claims.
A viewer who finishes a five-claim video has been given more information and less reason to buy. Nothing has been established firmly enough to survive the next scroll.
Why Repetition Does the Work
Robert Zajonc named the mechanism in 1968. His experiments found that people develop a preference for things simply from having encountered them before, across objects, sounds, and even nonsense words. The effect held for stimuli shown for milliseconds, below the threshold of conscious awareness. Familiarity produces liking without the viewer knowing why.
Herbert Krugman added the sequencing in 1972. His three-exposure theory describes what each encounter does. The first triggers a "what is it" response. The second prompts "what of it," where relevance gets evaluated. The third functions as a reminder and a decision anchor.
Read those two together and the arithmetic becomes uncomfortable. If a single message needs three encounters before it anchors, then a video carrying five messages has given each of them roughly one fifth of a single exposure. The audience has not received five messages. They have received noise five times.
The Layer Most Brands Repeat Is the Wrong One
Most content calendars hold the execution constant and vary the message. Same format, same shot list, same creator style, same opening frame, and a different claim each week. Monday is about durability, Wednesday is about price, Friday is about the new colourway.
The principle requires the exact opposite. Hold the message fixed and vary the execution until the variation runs out.
Take the plug-in device. The message is ease of use. That single idea supports a long list of separate executions: a parent setting it up in a child's bedroom before bedtime, a demonstration against the mess of an old coil, a five-second hands-only install, a guest-room setup before family arrives, a rental apartment where nothing can be mounted, a comparison of what the nightstand looked like before and after.
Six pieces of content. One message. Nobody feels repeated at, because the surface changed every time.
Run the inverse and you produce six pieces that each said something different, using the same visual grammar, and the audience remembers the format rather than the point.
Why This Fails Inside Organisations
The creative logic here is not hard to grasp. Brand teams understand it in the room and abandon it in the brief, and the reason is rarely creative.
Locking one message means telling people no. Repeatedly.
The product manager who fought for the 45-day refill wants the 45-day refill mentioned. The regulatory team that cleared specific efficacy wording wants that wording used. The commercial lead who negotiated the bundle price wants the bundle price visible. Each request is individually reasonable and each one is a claim.
What arrives at the creator is a brief with five priorities, which is a brief with none. The creator does the honest thing and mentions all five, because that is what was asked. The resulting video is competent, accurate, thorough, and forgettable.
This is why message discipline is a governance problem wearing a creative costume. The person who can enforce it is not the copywriter. It is whoever is senior enough to tell three internal stakeholders that their feature goes in the supporting slot this quarter.
The Measurement Trap
The second failure is structural. Message consistency pays off across months. Content dashboards report weekly.
In week three of a locked-message campaign, the numbers look flat. Nothing has compounded yet, and the content feels repetitive to the team producing it, because they have seen every asset a dozen times and the audience has seen one or two.
That gap is where most single-message campaigns die. Somebody proposes freshening the angle, the brief widens, and the program returns to saying a hundred things once. The abandonment usually happens right before the mechanism would have started working.
Where "A Hundred Times" Stops Being True
The heuristic is directionally right and numerically wrong, and it is worth being honest about that.
Effective frequency research has generally found gains flattening sharply past roughly ten exposures within a purchase cycle. Beyond that, the same spend produces diminishing rather than compounding returns.
Creative wear-out is messier still. A recent analysis of 2.8 billion ad impressions found the effect is highly heterogeneous: four campaigns out of thirty showed significant wear-out after one or two exposures, while ten showed almost none after fifty. The variance tracked to creative complexity, category involvement, and whether multiple executions were in rotation.
That last variable is the one worth holding onto. Rotation of execution is what buys you frequency without decay. The campaigns that survived fifty exposures were not repeating one asset fifty times. They were repeating one idea across many assets.
So "a hundred times" is not a media plan. It is an argument about which layer stays fixed.
Running It in Practice
Write the message as one sentence, then delete the second sentence. If the core idea needs two sentences to state, it is two ideas and you have not chosen yet. The plug-in device message was four words long.
Build the scenario list before the content calendar. One message multiplied by twelve genuine use situations gives you a quarter of content with no repetition fatigue. Starting from a calendar and filling slots produces the opposite: twelve messages and one format.
Put supporting features in a named slot, at the end. Other benefits are not banned. They are positioned. Giving stakeholders a defined place for their feature is what makes the internal conversation survivable, and it costs almost nothing because the closing seconds were rarely doing persuasive work anyway.
Set the review window before you start. Agree in advance that the message stays locked for a full quarter regardless of week-three numbers. The decision to hold has to be made when nobody is anxious, because it will be questioned when somebody is.
Rotate execution, and count it. If your last ten pieces used three formats, your effective frequency ceiling is lower than you think. Track distinct executions per message the way you track spend.
The Uncomfortable Arithmetic
Every additional claim in a piece of content divides the attention available to the one that matters. Five claims do not give the audience five chances to be persuaded. They give one message a fifth of a chance.
The Rp 119 million video was not the most informative asset in that dataset. Several failures were more thorough, better lit, and more complete. It was the one that had decided what it was about.
For the video-level mechanics behind that finding, see the analysis of 5,871 FMCG affiliate videos. For the briefing fields that carry a locked message through to production, see the nine brief fields that decide affiliate revenue.
Credit where it is owed: the framing of this piece comes from a short video by Indonesian marketer @fadelyulian, who put it more cleanly than we have. Say one thing a hundred times, not a hundred things once.
