Finding a product trend before it peaks is a timing problem. Most sellers and creators notice a product after it is already loud on one platform. The fix is not a hotter take. It is a multi-source sequence that treats each platform as a different stage of the same demand.
Google Trends alone is a late instrument for many SKUs. It is relative, it is Google-shaped, and it often ignores names that are already moving on TikTok or in Amazon search. TikTok-only dashboards miss purchase intent. Amazon-only suites miss cultural ignition. The workflow below is a lifecycle map, not a screenshot of today's viral list.
Trends MCP is one way to pull those sources as normalized series through MCP or REST. The same logic can be executed with more tab-switching if a team prefers manual charts. The method is the durable part.
Why single-platform tools miss early products
Each channel has a different job in the consumer journey:
- TikTok hashtag and topic volume often reflect cultural awareness first.
- YouTube search often reflects "how does this work" and review intent.
- Amazon search reflects purchase-shaped queries.
- Google Search reflects mainstream information seeking.
- Google Shopping reflects active price comparison, usually later.
A product rising on TikTok, quiet on Google Search, and only just appearing on YouTube is often early. A product flat on TikTok, rising on Google Search, and spiking on Shopping is often near or past a comfortable entry for new sellers. Category exceptions exist: some tech accessories are Amazon-first; some fitness ideas are YouTube-first. The map is a prior, not a law.
For source-level context see TikTok trends and Amazon search trends. For tool-by-tool dropshipping stacks see best tools for dropshipping product research.
Ranked tools for this workflow
Ranked by demand history and multi-source access, not by who has the prettiest winning-product gallery.
| Tool | Role | Cost notes |
|---|---|---|
| Trends MCP | Cross-source growth and history for the same keyword | Free: 100 requests/month. Starter: $19/month. MCP and REST |
| Google Trends | Free Search confirmation | Free |
| Amazon research suites (for example Jungle Scout) | Unit-sales style validation | Commonly listed from about $29/month; confirm current pricing |
| Ad intelligence (for example Minea) | Creative and auction reality | Commonly listed from about €49/month; confirm current pricing |
Trends MCP does not replace sales estimators or ad libraries. It is the comparison layer.
Step 1: Start with TikTok trajectory
Look for topic or hashtag series that have moved from a low normalized band to a clearly higher band over several weeks without a matching Google or Amazon spike. That gap is the discovery window.
get_trends(keyword='example product', source='tiktok', data_mode='weekly')
get_trends(keyword='example product', source='google search', data_mode='weekly')
If TikTok is elevated and Search is still low, treat the product as early-awareness, not as proven revenue.
Filter out memes, sports, and celebrity names unless the business is media. Keep nouns that can sit on a product page.
Step 2: Cross-reference Amazon search
Amazon search is closer to buy intent than a For You clip. TikTok up and Amazon still flat is awareness. Amazon 3-month growth outpacing Google Search 3-month growth is often the purchase-intent acceleration phase, still before mass Shopping comparison in many categories.
get_growth(keyword='example product', source='amazon', percent_growth=['3M', '6M'])
get_growth(keyword='example product', source='google search', percent_growth=['3M', '6M'])
Do not skip supplier and review-quality checks. A rising Amazon query can still be a junk category.
Step 3: Split Google Shopping from Google Search
Shopping captures price-comparison behavior. Search can rise on curiosity while Shopping stays quiet. When Shopping growth starts to beat Search growth, competitive pricing pressure often follows. That is a late entry for thin-margin dropshipping more often than a beginning.
get_growth(keyword='example product', source='google shopping', percent_growth=['3M'])
get_growth(keyword='example product', source='google search', percent_growth=['3M'])
A pattern that is still useful years later: Search clearly up, Shopping only slightly up, is mid-cycle. Both hot is crowded.
Step 4: Validate with YouTube
YouTube search marks the tutorial and review phase. It often sits after the first TikTok wave and before Search is fully mainstream. Rapid YouTube growth after several weeks of elevated TikTok is a mid-stage tell.
get_growth(keyword='example product', source='youtube', percent_growth=['3M'])
If YouTube never appears, the trend may stay entertainment. That can still be a content opportunity. It is a weaker wholesale bet.
Step 5: Place the product on a lifecycle map
Use this as a worksheet, not as a live leaderboard.
| Stage | TikTok | YouTube | Amazon | Google Search | Google Shopping |
|---|---|---|---|---|---|
| Pre-trend | Low | Low | Low | Low | Low |
| Early | Rising | Low | Low | Low | Low |
| Mid | High | Rising | Rising | Rising | Low |
| Peak | High | High | High | Peak | Rising |
| Saturation | Declining | Mixed | Mixed | Declining | High |
The usual comfortable entry for new sellers is mid: TikTok has been elevated for weeks, Amazon and Search are turning up, Shopping is not yet the whole story. Peak and saturation are for brands that already have supply and reviews.
Trends MCP can compute the position by comparing normalized scores across sources for one keyword. The free tier is 100 requests per month, so batch the shortlist. Starter is $19 per month when the catalog of candidates is large.
Systematic discovery without a "today's winners" page
Rather than hunting celebrity SKUs, rank TikTok by growth or latest value, then keep names where TikTok is high and Google Search is still low. That gap list is the working set.
get_ranked_trends(source='tiktok', sort='wow_pct_change', limit=50)
Human review still removes non-products. Then run Amazon and Shopping only on survivors. Quota lasts longer that way.
Common mistakes
Google Trends only. Relative interest hides small but profitable niches and lags some social names.
Awareness as revenue. Rising Search is not sales. Amazon and Shopping are better intent proxies.
One universal lag. Beauty is often TikTok-first. Some accessories are Amazon-first. Fitness may be YouTube-then-Google. Learn the category.
One-week spikes. Three to six weeks with a second platform is a sturdier bar than a single viral day.
Skipping unit economics. Trend heat does not pay returns, ads, or broken widgets. A rising series is a hypothesis to source and test with small inventory or a landing page, not a purchase order for ten thousand units.
Ignoring category-specific lag. Apply the five-source map, then rewrite it after a month of notes. If Amazon consistently leads TikTok in a niche, stop waiting for a hashtag that will never come. If Shopping never moves, the category may be gift-and-impulse rather than comparison shopping.
A written one-page playbook per category (which source leads, how many weeks to wait, when to kill a candidate) is more valuable than another tool login.
FAQ
Can this workflow run in a spreadsheet?
Yes. Export or copy series weekly. MCP is faster when the team already works in an assistant. REST fits notebooks.
Does Trends MCP give Amazon unit sales?
No. It gives search-style demand series for Amazon as a source, plus other platforms. Sales estimators remain a separate purchase.
Where should a beginner start?
One TikTok rank pull, ten product-like names, Amazon plus Search growth on those ten. Ignore the rest until that habit exists.