Six of the eight highest-reach failures in this dataset came from one product category.
Haircare. Videos with 107,968 views, 51,188 views, 25,040 views, all selling nothing. Meanwhile a roach spray video reached 1,239 people and sold eight units.
We saw that split repeatedly while auditing 5,871 affiliate videos across nine FMCG product lines in Indonesia. The zero-revenue rate ran from 16.6% on one line to 86.5% on another, using the same seller, the same affiliate pool, and the same platform mechanics.
The gap is not execution quality. It is what the buyer was doing when the video found them.
Two Kinds of Product
Some FMCG purchases begin with a problem the buyer already has. Rats in the kitchen. Mosquitoes at night. A blocked drain. The buyer is not browsing. They have an active irritation and they are looking for the thing that stops it.
Other purchases begin with exposure. Nobody wakes up urgently needing a reed diffuser. The category gets bought when someone sees one, likes it, and decides their living room could use it.
Affiliate video works differently in each case, and the failure rates show it.
| Product line | Purchase trigger | Videos | Zero-revenue rate |
|---|---|---|---|
| Plug-in mosquito repellent | Active problem | 824 | 16.6% |
| Pest adhesive | Active problem | 276 | 29.3% |
| Drain cleaner | Active problem | 184 | 60.3% |
| Home air device | Mixed | 643 | 64.7% |
| Shampoo | Discovery | 1,546 | 69.3% |
| Car air freshener | Discovery | 733 | 73.7% |
| Roach spray | Active problem | 326 | 75.8% |
| Reed diffuser | Discovery | 1,043 | 82.3% |
| Baby wipes | Trust-gated | 296 | 86.5% |
The ordering is not clean, and the roach spray line sits awkwardly at 75.8% despite being an obvious problem-solver. More on that below. But the two ends of the table hold: the lines built on an active, urgent, physical problem convert best, and the lines built on aesthetic preference convert worst.
Why Discovery Categories Get Reach and Lose Money
Haircare and home fragrance produced the highest view counts in the entire dataset. A shampoo video hit 1,555,338 views. A reed diffuser video hit 1,330,998.
Both are watchable categories. Before-and-after hair transformations travel. So do aesthetic room shots. The content performs as content, which means TikTok distributes it beyond the people who might buy.
The reed diffuser video with 1.3 million views sold 163 units, converting at 0.12 per thousand. The roach spray video with 1,239 views sold 8, converting at 6.46 per thousand. Fifty-four times better conversion on a thousandth of the reach.
Discovery categories will give you reach whether or not the video sells anything. That makes them satisfying to run and difficult to evaluate.
Why Problem Categories Convert on Almost No Reach
Nobody watches a drain cleaner video for entertainment. The 720-view drain cleaner video that sold seven units reached almost exactly the population it needed: people whose drain is blocked right now.
When the audience arrives with an active problem, the video's job shrinks to three questions. Does this fix my thing. How do I use it. Where do I buy it. There is no persuasion phase because the intent already exists.
That is why the top-performing videos in the pest control lines were product-first and unglamorous: device in frame one, mechanism stated, price named, cart location given. They read as instructions rather than content.
The Roach Spray Exception
Roach spray should behave like the other pest lines and does not. 75.8% of its videos sold nothing, sitting closer to the discovery categories than to its own product family.
The likeliest explanation is competitive rather than behavioural. Insecticide spray is the most saturated shelf in Indonesian pest control, with entrenched incumbents and near-universal household familiarity. A viewer with a roach problem already knows what to buy. The affiliate video is not answering "what solves this," it is arguing "switch brands," which is a much harder ask on a Rp 30,000 product.
The plug-in device line at 16.6% is the mirror image: a newer format, less familiar, where the video genuinely is the buyer's introduction to the mechanism.
Category difficulty is not only about urgency. It is also about whether the viewer needs to be told something they do not already know.
Baby Wipes and the Trust Gate
Baby care failed at 86.5%, the worst rate in the sample, and it does not fit either bucket cleanly.
Parents buying wipes for a newborn have an active need, so the intent is there. What is missing is permission. The purchase is gated on safety, ingredients, and whether the recommendation source is credible. A creator holding up a pack and naming a price does not clear that gate.
The top-performing baby care videos in the sample were also the smallest in absolute terms, topping out at Rp 486,430 against Rp 119 million on the mosquito line. The category can be sold, but not with the mechanics that work on pest control.
What to Change Per Category
For active-problem lines, optimise for specificity over reach. Narrow targeting, problem named in the first two seconds, mechanism shown, exact price. Stop treating low view counts as underperformance when conversion is strong. The 1,239-view video that sold eight units is doing its job.
For discovery lines, accept that most views will not convert and set the economics accordingly. Judge these videos on sales per thousand views rather than total views, and expect the ratio to sit an order of magnitude below the problem categories. Budget for volume, because you are paying for the small fraction of an unqualified audience that happens to be in market.
For trust-gated lines, the constraint is credibility rather than clarity. Certification visible on pack, ingredient specificity, and creator category authority matter more than offer mechanics. A generalist creator holding a baby product is the wrong instrument.
For saturated lines, the video has to give the viewer a reason to switch, not a reason to buy. That means a named point of difference against what they currently use, which is a different script from the one that introduces a new format.
The Planning Error
Most affiliate programs run one brief template across the portfolio, then read the results as creator performance. A pest control brief applied to baby wipes produces a competent video that sells nothing, and the creator takes the blame.
The nine lines here were run by the same seller through the same affiliate pool. The five-fold spread in failure rate was set before any creator was briefed, by which products went into the program and what the video was being asked to do for each one.
For the video mechanics that separate winners from failures inside a category, see the analysis of 5,871 FMCG affiliate videos. For why reach and conversion diverge in discovery categories, see the reach-conversion tradeoff.
The Short Version
Ask what the buyer was doing thirty seconds before your video reached them. If they had a problem, sell the mechanism. If they were scrolling, expect to pay for volume. If they were worried, sell credibility first.
Running the same playbook across all three is how a program ends up with an 86% failure rate on one line and 16% on another while blaming the creators for both.
