Most product trends are catchable about three weeks before the wave crests. Past the peak, you are competing with 200 sellers selling the same thing and the math no longer works. The skill is recognising the signals early. Below are the specific signals I watch, the kill rule that prevents catching a falling knife, and the way to act once you spot one.
The four signals
Signal 1 - TikTok organic momentum without paid amplification. When a product starts generating 50K+ organic views on multiple non-sponsored videos within a 7-day window, the wave is starting. The viewers will start searching the product within 1-3 weeks.
Signal 2 - Reddit threads in niche subs starting to discuss a specific product type. Reddit lags TikTok by about a week. When threads with 100+ comments appear in /r/[niche] asking "where can I buy" a specific item, the search-volume wave is already building.
Signal 3 - Google Trends curve turning upward sharply. Set up Google Trends alerts for the keywords you watch. A 50%+ week-over-week climb is the signal. A 200%+ climb means you are now in the public-awareness phase and you need to move fast.
Signal 4 - new Etsy or Amazon listings appearing for the product type, faster than usual. Use Helium 10's New Listings tool or just manually check. When 20+ new listings appear in a week for a category that was static, sellers are racing in.
One signal alone is suggestive. Two together is a real signal. Three or four together is a trend in motion.
The timeline of a trend
Week minus 4 to minus 2 - signals start emerging on TikTok and Reddit. Only the early sellers see them.
Week minus 1 to 0 - Google Trends turns up. Listings start appearing. Early sellers are launching.
Week 0 to 4 - the wave peaks. Search volume is highest. Competition is high but margins are still good if you got in early.
Week 4 to 12 - the wave starts declining. New sellers continue piling in, attracted by the visible sales of the early movers. Margins compress.
Week 12 onwards - the trend is over. Sellers who entered late lose money. The few branded operators who built durable demand around the product type continue selling but at lower volume than peak.
The window to enter is week minus 2 to week 0. The window to exit is week 4 to 6.
The kill rule
If by week 0 you have not committed to a product, do not chase. The math will be worse than the previous trend, and you are now competing with sellers who have a 3-4 week head start.
Falling-knife products look like opportunities. They are not. They are graveyards.
Better to miss a trend than to enter late on one. Three to five trends a year are catchable. Missing one is not a problem because another is forming somewhere else.
The wave is not the prize. The discipline to spot the right one and ignore the wrong ones is the prize.
How to act when you spot one
Fast. The whole point is speed.
Day 1 - source samples from US or fast Asian suppliers. Pay for express shipping. Skip the 30-day boat.
Day 2-5 - listing copy, photos (use AI if needed for speed), Shopify or marketplace setup.
Day 6-10 - launch with paid ads. Test 4-6 creatives. The winning creative usually emerges within 48 hours.
Day 11-30 - scale the winner. Reorder inventory immediately. Maintain the listing aggressively.
Day 30-60 - milk the wave. Watch the signals for decline. Be ready to wind down.
The cycle is 30-60 days from spot to wind-down. The sellers who catch trends well make most of their annual income in two or three of these cycles per year.
What does NOT count as a trend
One viral TikTok video. Not enough signal. Could be a one-week novelty.
Influencer-driven peaks. If the trend exists only because one Kardashian-tier celebrity wore it, the trend dies when the celebrity moves on. Different from organic momentum.
"Trends" you read about in Forbes or Bloomberg. By the time those publications cover it, the trend is past peak and probably declining.
Saturated categories with year-on-year baseline demand mistaken for "trending". A 5% climb in a category that already does $50M/month is normal noise, not a trend.
The role of branding
The biggest profit on a trend is captured by the operator who turns the trending product into a brand before the wave breaks. The wave brings the traffic. The brand captures the long-tail.
Artur's car towel started as a trending product. He could have run the wave for 60 days and stopped. Instead he branded the product, kept the listing alive past the wave, and is still making money on it because the brand has its own demand independent of the original trend.
Detail in how to brand a product from scratch.
How many trends to chase per year
Three is the practical cap for a one-person operation. Each trend is 30-60 days of intense work. More than three a year and you cannot give any of them the attention they need.
Better to catch three properly than to chase eight badly.
For the broader product research playbook, read how to do product research for Amazon and how to use AI to find winning products. The full product-trend module - including the signal monitoring setup - is in the course. Watch the four signals. Catch one wave. Brand the product. Repeat.
