In a recent group with women in their early twenties, I asked everyone to open their phones and show me the last skincare video they had saved. Seven phones, seven videos (one of them, awkwardly, a lip plumper). Not one had been searched for. Every single video had simply arrived in the feed, and most participants couldn't say who the creator was without checking (I suspect that's the point of the format). One woman scrolled back and found she had saved three different videos of the same toner, from three different creators, over about two weeks. She bought it on the third. "Everyone was using," she said, and then, after a pause, "or maybe only my phone thinks everyone was using."
That second sentence has stayed with me. Beauty counters, magazines, and older sisters used to be the places a young shopper met a new product, and each came with a face you could question. On TikTok the introduction is made by an algorithm that has no face, and the repetition it creates starts to feel like consensus. This post looks at how that loop turns a scroll into a purchase, which voices inside the feed actually carry trust, and what brands should research before they spend another dollar chasing virality.
Why the feed feels like advice
TikTok Shop was designed to collapse discovery and purchase into one place. Retail Asia's report on the Singapore launch describes product recommendations delivered through in-feed videos, livestreams, and a tab that lists a creator's products. So the recommendation, the demonstration, and the checkout sit a thumb's width apart. For a category where the result is visible on a face within seconds of application, that's a powerful arrangement.
But the feed also does something quieter. It learns what you linger on and gives you more of it, until a niche product looks universal. The clearest evidence I've seen of how fast this happens came from outside Singapore. The Straits Times, reporting an AFP story in November 2025, described a dermatologist at Northwestern University who posed on TikTok as a 13-year-old interested in beauty, and found the algorithm "saturated" her account. Across 100 videos from 82 profiles, "get ready with me" routines used six products on average, at an average combined cost of US$168, and the most-viewed videos contained an average of 11 potentially irritating active ingredients. Nobody asked those girls to buy six products. The feed simply showed them that everyone else had.
Is Singapore different? Probably not in the mechanism, and I suspect not much in the outcome either. Participants here describe the same experience of a product "following" them for a week. What differs is how they talk about it afterwards (which is where the research gets interesting).
The trust hierarchy inside a single app
My first reading of the transcripts was that TikTok had simply replaced the beauty counter as the trusted adviser. Actually, I'm overstating it. Participants trust the feed to introduce products, and they distrust almost everything else about it. They know sponsored content is common. They know the platform carries junk, and they're right to. MDDI's 2024 Online Safety Poll found that among respondents who encountered harmful content on designated social media services, 40 per cent cited TikTok. So young shoppers are sceptical by default. What they do instead is sort the voices in their feed, fast and fairly consistently.
At the top sits the creator whose skin looks like theirs, the same tone, the same oily T-zone, the same humidity (a big deal in a climate like ours). Participants call these creators "real" and they forgive a lot of sponsorship for that realness. Just below sits the explainer, the creator who reads ingredient lists aloud and tells you why niacinamide stings. Harper's Bazaar Singapore's July 2026 interview with Ingu Skin's founder shows how far that role can travel. A Thai creator who explains skincare science started posting in English, picked up followers in Singapore, and began selling his brand here in January 2026. At the bottom sits the haul, the unboxing and the twelve-product routine. Hauls create want. They don't seem to create belief, at least not on their own.
Traditional media hasn't disappeared from this hierarchy. It has moved to the end of the journey. More than one participant told us they check whether a product is stocked at Sephora or Watsons before buying, as a rough sign that "someone has checked it". That's a generous assumption (and not quite how it works). HSA explains that a cosmetic product notification does not mean the product is approved, and that dealers are directly responsible for their products' safety. The shelf feels like a filter, and the shopper lets it act like one.
From scroll to basket, in four moments
I tried at first to describe this as a funnel. It isn't one, or not a tidy one. People loop back, and the same product can sit at "want" for months before something tips it. What I kept hearing, though, were four moments that appear in roughly the same order, laid out below as the Feed-to-Basket Loop. I'm still testing whether the last step behaves differently for shoppers over thirty.
The Feed-to-Basket Loop
Ambush
The product arrives uninvited in the feed, usually in a haul or routine video.
Echo
It reappears from different creators within days, and repetition starts to read as consensus.
Mirror Check
The shopper searches for a creator with similar skin, and reads the comments for bad reactions.
Permission
A low price, a retail shelf, or a friend's approval turns want into a basket.
The Echo stage is where brands tend to over-invest, since it's the stage that paid seeding can buy most directly. The Mirror Check is where they under-invest (partly because it's harder to buy), and I think it's where most sales are really decided. A shopper who can't find anyone with her skin type using the product will often park it indefinitely, however many times it echoes. And Permission is where price does its work. The Ordinary, one of the brands the Northwestern study found disproportionately represented, sells on Sephora Singapore under the promise that "Quality Skincare Belongs to Everybody" (a price point that makes a TikTok impulse feel low-risk). A $15 serum needs far less permission than a $150 one (my numbers are illustrative, not prices from any shelf), which is the same logic we traced in when Singaporeans splurge and when they save on skincare.
What young shoppers say about TikTok, and what their shelves show
Ask directly and participants give answers that protect their judgement. Look at their saved videos and their bathroom shelves, and the picture shifts.
| What participants say | What their phones and shelves show |
|---|---|
| "I don't buy things because of TikTok" | Most recent skincare purchases can be traced back to a saved or liked video |
| "I can tell when it's sponsored" | They discount obvious ads, but rarely question a relatable creator who received free product |
| "I do my own research" | Research usually means searching TikTok itself and reading comments, not checking ingredient safety elsewhere |
| "Viral products are overhyped" | They own several, and keep the ones that passed their Mirror Check |
| "I trust dermatologists more" | True in principle, but the dermatologist they quote is often a creator they follow |
The sponsorship row matters for compliance as well as insight. The ASAS guidelines for interactive marketing and social media say a connection between endorser and marketer that could affect the credibility of an endorsement must be fully disclosed, noticeably and in clear language. Formal complaints are few, though. In a February 2026 parliamentary answer, MDDI said ASAS received between 40 and 48 pieces of feedback a year from 2021 to 2025 about misleading social media ads, and that around 2 per cent involved undisclosed paid influencers or sponsored posts. My reading is that young viewers rarely complain about sponsorship they didn't notice. And the sponsorship they did notice, they simply scroll past.
Studying a discovery channel nobody admits to using
Ask Gen Z shoppers where they discovered a product and they give the answer that makes them look discerning. The phone is more candid, so in focus groups we begin there, asking participants to show saved videos, search histories inside the app, and screenshots they sent to friends. The evidence is on the screen, and it makes the conversation more honest (and a lot more fun). It's an extension of the approach we described in what Gen Z skincare shoppers actually do.
Where did the buying actually happen? "Scroll back to the first video you saw of the last product you bought. Now show me the video that finally made you buy it. What changed between the two?" The gap between the first and last video usually tells you which stage of the loop the brand won, and which one it nearly lost.
The bad-reaction stories come out better one to one, which is why we pair groups with in-depth interviews. So do stories (often told with a laugh) about buying something because a creator's skin looked "exactly like mine", which participants find a little embarrassing to say in front of peers. The same group dynamics shaped what we found in why Singaporeans trust one anti-acne serum over another, and they echo the social pressure we explored in whether Singaporeans want perfection or connection from beauty. The feed isn't unique to skincare, either. Restaurants went through their own version of this, which we covered in how Instagram changed restaurant choice in Singapore. And for brands entering from abroad, the loop can move faster than distribution, which is part of why K-beauty dominates Singapore.
The algorithm introduces, and the shopper still decides
TikTok is very good at the first two stages of the loop, and you probably can't opt out of them. Or rather, you can, but your competitors won't. But the sale is usually made at the Mirror Check and at Permission, by a creator who looks like the buyer and by a price or a shelf that makes the risk feel small. Brands that pay only for Echo may be buying repetition that never reaches a basket. I don't know how durable this is (the platform changes faster than any research cycle). What I keep coming back to is that woman with three saved videos of one toner, wondering aloud whether everyone was really using it, or only her phone.
What skincare brands ask about TikTok and product discovery
How has TikTok changed skincare discovery in Singapore?
Which TikTok creators do young skincare shoppers trust most?
Do skincare influencers in Singapore have to disclose sponsorship?
Does going viral on TikTok guarantee skincare sales?
How do you research TikTok influence on skincare purchases?
Finding out whether your skincare brand is losing sales at the echo or the mirror check
A product can go viral on TikTok and still stall before the basket, because young shoppers buy only after they see it work on someone like them. We run focus groups and in-depth interviews built around participants' own feeds to show where your brand wins or drops out of the loop.
Request a quote →
