Quick Answer
No single number tells you whether a skincare trend will last. Record six free public signals on the same day, the same way: US search interest, Korean search interest, retailer search results, new launches, human studies and regulatory notices. If five or six point up, build. If three or four do, watch and check again in three months. If two or fewer do, skip the trend or rework the idea.
A trend that looks unstoppable on your feed today has to still be selling on the day your product ships. With a Korean ODM, that day is months away. Our launch timeline guide puts a new formula at roughly 18 to 26 weeks from signed brief to a pallet leaving Korea, plus US import handling.
So the useful question is not whether a trend is hot, but whether it will still be rising two quarters from now. The method below answers that with six free signals and a simple count, instead of one viral video.
Key Takeaways
- A fad is a spike in attention that fades before most products built for it reach the shelf. A lasting trend keeps rising, or holds a higher floor, across several readings.
- Google Trends and Naver DataLab show relative interest on a 0 to 100 scale, not search counts. Read the shape of the curve.
- Log all six signals on one day and repeat every three months. Five or six rising means build, three or four means watch, two or fewer means skip.
- Studies and regulators move slowly. A fresh warning letter or restriction outweighs a viral month.
Why Does a Trend's Lifespan Matter More Than Its Peak?
A trend's lifespan matters more than its peak because your product reaches shelves months after you commit to it. If interest peaks while you are still approving samples, you launch into a falling curve with inventory you have already paid for.
A fad refers to a burst of attention that fades within a season or two, often set off by one video, one celebrity or one hero product. A lasting trend is a shift that keeps showing up in search, retail, science and regulation over several years, even when its curve dips between waves.
Fad vs Lasting Trend: How the Six Signals Compare
A fad and a lasting trend separate once you read the same six signals side by side. The table shows each pattern and where to check it for free.
| Signal | Free place to check | Fad pattern | Lasting pattern |
|---|---|---|---|
| 1. US search interest | Google Trends, United States, past five years | One sharp spike, then back near the old baseline | Rises in steps or seasonal waves, each low higher than the last |
| 2. Korean search interest | Naver DataLab search trend | Korean curve already falling while US interest climbs | Korean curve still rising or holding at a higher level |
| 3. Retailer search results | One US retailer's site search, logged each quarter | Result count flat or shrinking between readings | Result count growing on the same site |
| 4. New launches | Brand launch press releases and Korean supplier news, last 12 months | Launches bunched into one short window, then silence | Steady launches all year, and suppliers add new grades or formats |
| 5. Human evidence | PubMed, ingredient name, human study filters | Few or no human studies, mostly lab or animal work | Human studies appearing year after year |
| 6. Regulatory notices | FDA warning letters, MFDS notices, EU SCCS opinions | Fresh warnings, restrictions or claim crackdowns | No new adverse notices, or rules that clarify safe use |
No single row decides the call. A fad can have a real study behind it, and a lasting trend can have a flat quarter, so read the pattern across all six.
How Do You Read Google Trends and Naver DataLab Without Fooling Yourself?
Read both tools as shapes, not volumes. Google Trends divides each data point by the total searches for that place and time. It then scales the result from 0 to 100, so a value shows relative popularity, not a search count (Google Trends FAQ).
Google's training material adds that 100 marks the chart's peak. A falling line means relative popularity is shrinking, not necessarily the number of searches (Google News Initiative lesson).
Three habits keep the reading honest:
- Pull five years for the United States, not 90 days. A short window turns one viral week into a mountain.
- Add a steady benchmark term, such as moisturizer, to the same query. Google suggests comparing a term with common searches to understand its scale.
- Compare numbers only within one query pulled on one day. Each query is scaled to its own peak, so a 60 from March and a 60 from June are not the same thing.
Mind the noise warning too. Google says low-volume terms show as 0, and that statistical noise can create one-off spikes on queries with little search interest. A jagged line near zero is not a trend yet.
Naver DataLab works the same way for Korea. Naver Cloud's API documentation describes each value as a relative ratio, with the largest value in the results set to 100 (Naver Cloud docs). Run the Korean name of the ingredient or format and compare the shape of the two curves.
A curve already falling in Korea while it climbs in the US calls for caution; rising in both is a stronger sign. Our 2027 ingredient trend forecast explains why Korean ODM pipelines tend to surface in Korea first. Use that as context, not as a countdown.
Retail Shelves and New Launches Show Whether Money Is Following
Search data shows curiosity. Retail and launch data show whether brands and buyers are putting money behind the trend, which is what keeps it alive after the first wave.
Signal 3 is retailer search results. Type the trend term into one US retailer's search box and note the result count and date. Repeat on the same site three months later. Sites count results differently, so never compare counts across retailers.
Signal 4 is new entrants over the last 12 months. Search brand launch press releases for the term and scan Korean ODM and supplier news for new raw materials built around it. Launches spread across the year, with suppliers adding new grades, point to a market that is still building. Launches bunched into one season, then silence, often mark the top.
Export data is too broad to judge one trend; our K-beauty export data breakdown covers that bigger picture.
"I'm Liz, I run altameet from Manhattan, NYC. Given the choice, I would back a trend that is boring and still climbing over one that is exciting and already at its peak. If you want a quick gut-check on whether Korean ODM fits your project, I'll give you 15 minutes free."
What Do Studies and Regulators Add That Search Data Can't?
Studies and regulators move slowly, and that is exactly why they help. A trend with human studies piling up and no regulatory trouble has a base that a viral month does not.
Signal 5 is evidence depth. Search PubMed for the ingredient, filter to human studies and clinical trial article types, then open the results-by-year timeline. Are human studies appearing year after year, or is the record mostly lab and animal work? PubMed's user guide notes the timeline counts every publication date a record has, print and electronic. Its total can differ from the result count (PubMed User Guide). Read the slope, not the sum.
Signal 6 is regulatory notices. Check three places: FDA warning letters for claims or products tied to the trend, notices from Korea's MFDS, and opinions from the EU Scientific Committee on Consumer Safety. A fresh warning letter, a new restriction or a draft opinion questioning safety is a reason to slow down. A stable rulebook is neutral.
These two signals also shape your copy. When human evidence is thin, keep claims to how the product looks and feels.
Score the Signals: Go, Watch or Skip
Give each signal a direction on the day you check it: rising, flat or falling. Then count how many of the six are rising and apply a three-step rule.
| Signals rising | Call | What to do next |
|---|---|---|
| Five or six | Go | Brief your ODM and build the trend into the product from the start. |
| Three or four | Watch | Check again in three months. Develop a version that still works if the trend cools. |
| Two or fewer | Skip or rework | Drop the trend, or keep the format and change the hook. |
Write one log line per signal: date, tool, query, value and direction. A shared spreadsheet is enough, and the second reading is what shows direction.
Two tie-breakers help when the count lands in the middle. A new warning letter or restriction moves the call down a step on its own. A trend that fits your customer and price point can move from watch to go after one more reading.
Paid tools, such as social listening platforms and sales panels from Circana or NielsenIQ, go deeper; this method sticks to free sources. Whether the US market has room for a new brand at all is a separate question, covered in our skincare market saturation guide.
What Does Betting on a Fad Cost a New Brand?
Betting on a fad costs everything you commit before the first sale: development, a full minimum order, packaging tooling and launch marketing. Those costs land months before you learn whether the trend held.
You can lower that exposure without avoiding trends altogether:
- Put the trend in a variant, not the base. Build a stable base formula and treat the trend ingredient or texture as one version of it.
- Choose packaging you can reuse. Stock bottles and jars outlive a trend. Custom molds tied to one look do not.
- Keep claims modest. If the trend fades, a product sold on its feel and results still has a reason to exist.
- Book the second reading. Put the next check on the calendar before you sign the purchase order.
A trend that passes five or six signals is worth building around; one that passes two or fewer is usually a better story than a product. To place a trend in a first lineup, read our guide to choosing a first collection, then the US launch playbook for Korean cosmetics.
Trend Signal Log Checklist
Get the six-signal log and scoring sheet as a free PDF, straight to your inbox.
Frequently Asked Questions
How long should I track a trend before I commit to a product?
Track it for at least two readings, three months apart, before you brief a manufacturer. One reading shows a level; only the second shows direction. If your calendar is tight, commit to a stable base formula now and add the trend element once the second reading confirms it.
Do TikTok views prove a skincare trend will last?
No. Views measure attention on one platform, and attention moves on quickly. Treat a viral clip as the reason to run the six signals, not as one of them. If search, retail and launch data do not follow, the trend has not reached buyers yet.
Is a trend that started in Korea safe to bring to the US?
Not automatically. A Korean head start is useful context, but the US search curve, US retailer data and US rules decide your launch. Confirm that your planned claim is allowed for a cosmetic in the US before you build a product around it.
What if the six signals disagree with each other?
Mixed signals mean watch, not go. Note which ones disagree. Rising search with no new launches points to early demand, while a wave of launches with falling search points to brands arriving late. Check again in three months and decide on the second reading.
By Liz Song, K-beauty sourcing consultant.
