Disclosure: this post is by SkyData. We built and own the three paid Apify Actors used below (pay per result). You don't need them to follow this method: you can do a rougher version by hand, for free, on the public websites: search AliExpress and sort by orders (the site shows rounded counts like "10,000+ sold"), compare the keywords in Google Trends, and look the brands up in Google's Ads Transparency Center. The Actors make it faster and repeatable, and add the unrounded counts, per-search reports and ad statistics shown below.
Before buying stock or building a store around a product, three questions save a lot of money:
- Does anything like it sell at all? (AliExpress sold counts)
- Is interest growing or fading? (Google Trends)
- Who is already advertising it? (Google Ads Transparency Center)
Each answer is weak on its own. Together they make a quick first filter. Here is the whole workflow on three room-lighting ideas: LED strip lights, sunset lamps and galaxy projectors. All numbers are from runs on 3 Nov 2026.
Step 1: What sells on AliExpress?
For each keyword I asked AliExpress to sort by most orders and kept the first 60 organic products (ads excluded).
| Search | Listings AliExpress reports | Median price today | Median sold per product (cumulative) | Products with 1,000+ sold |
|---|---|---|---|---|
| LED strip lights | ~47,000 | $9.93 | 12,878 | 59 of 60 |
| Galaxy projector | ~11,000 | $12.05 | 2,768 | 60 of 60 |
| Sunset lamp | ~9,500 | $7.04 | 545 | 25 of 60 |
How to read this, and how not to:
- It is a biased sample on purpose. Sorting by most orders puts the best sellers first, so these 60 products are the top of each niche, not a typical product. The medians describe the best sellers, and you cannot use them to estimate total market demand or what an average new listing would sell.
- The sold counts are cumulative totals reported by AliExpress since each listing went live. We don't verify them, and they are not recent sales: a listing from 2023 with 80,000 sold may sell very little today.
- The "listings" figure is AliExpress's own estimate and changes from visit to visit; treat it as a rough size.
- Look twice at unusual listing dates. One car-light listing showed a listing date 63 days earlier and 79,054 sold (about 1,255 a day). That doesn't prove anything is wrong (the date may have been reset by an update or re-listing, or the product may really sell that fast), but check the product page before you trust it.
Results also move between runs: an hour later, the same search gave a median of 35,762 sold for LED strip lights instead of 12,878, because AliExpress returned a different set of best sellers. Compare several runs before deciding anything.
To measure recent sales you need two runs: run the same search a day or a week later and compare the count per product (soldSinceLastRun).
Step 2: Is interest growing or fading?
Same three keywords in one Google Trends comparison, last 5 years, once for the US and once worldwide. "Change" compares the average of the last quarter of the period with the first quarter.
| Keyword | US: average (0–100) | US: change | Worldwide: average (0–100) | Worldwide: change |
|---|---|---|---|---|
| LED strip lights | 33.7 | +1% (stable) | 48.4 | +16.7% (stable) |
| Galaxy projector | 8.7 | +41.8% (rising) | 16.2 | +68.5% (rising) |
| Sunset lamp | 6.7 | −32.4% (falling) | 10.2 | −19.5% (stable) |
Google Trends numbers are relative: within one comparison, 100 is the busiest week of the busiest keyword, and everything else is scaled to it. So you can say "in the US, LED strip lights get about five times the search interest of sunset lamps", but you can't compare the US column with the worldwide column (each is normalized separately), and none of these numbers are search volumes.
All three are strongest in November–December (gift season); LED strips and galaxy projectors also peak around April. Google's "breakout" related searches for galaxy projectors were mostly brand names (Pococo, Govee, West & Arrow): a hint that branded sellers already capture part of that interest.
Step 3: Who is already advertising it?
The Google Ads Transparency Center is Google's public ad library. I checked two brands from this space.
- govee.com: Google shows about 1,000–2,000 ads in the last 30 days (4,000–5,000 in total). Of the 100 most recent ads: 15 were new in the last 30 days (8 in the last week) and 76 were still live after 60+ days. Long-running ads are worth a look because the advertiser chose to keep them, but duration alone doesn't show that an ad is profitable: the library has no spend, clicks or sales. Ads appeared in all 8 countries checked (US, GB, CA, AU, IN, DE, FR, BR).
- blisslights.com: No matching ads found under the selected search filters (this domain, last 30 days, all regions).
More limits: for big advertisers I analysed the 100 most recent ads, not all of them, and a domain search can include ads from other advertisers that point to the same site (for example resellers), so check the advertiser names.
What I'd conclude (and what I wouldn't)
- Sunset lamps: lowest sold counts among the best sellers, falling US interest, and one product accounts for 43% of all units sold in the sample. Weakest of the three.
- Galaxy projectors: interest is rising and every top-60 product shows 1,000+ sold, but brand names dominate the breakout searches. Worth a closer look, with a clear angle against the brands.
- LED strip lights: very large proven demand at the top of the market and at least one big, active advertiser (Govee). Hard to stand out on price alone.
None of this proves a product will sell for you. It tells you where to dig deeper: track sold counts weekly, read reviews, check shipping times and margins.
Cost and time
The three runs took 82 s, 78 s and 28 s. At the Actors' Free-plan list prices (as on their Pricing tabs on 3 Nov 2026):
| Step | What is charged | Cost |
|---|---|---|
| AliExpress | 180 products × $0.0015 + 3 search reports × $0.01 + Actor start $0.00005 | $0.30 |
| Google Trends | 2 reports × $0.003 + Actor start $0.00005 | $0.006 |
| Ads Transparency | 1 competitor report × $0.02 + 100 ads × $0.001 + Actor start $0.00005 (blisslights had no matching ads → not charged) | $0.12 |
| Total | ≈ $0.43 |
Each run also has Apify's Actor start event: $0.00005 for a run with up to 1 GB of memory, plus $0.00005 for each additional GB (these runs use 256 MB, so one $0.00005 fee each). Paid Apify plans get 5–15% off the event prices. On the Free plan, these runs are paid from Apify's monthly free usage credit; under Apify's rules, runs by Free-plan users don't generate revenue for us as the developer.
Run it yourself
The Python script that runs all three steps with the official apify-client is on GitHub: github.com/Scream3000/skydata-product-research-example. Before anything runs it shows the most the runs can cost, asks you to confirm, and caps every run (default limit for the whole workflow: $1.00).
Get your token from console.apify.com → Settings → API & Integrations, and never commit it.
Linux / macOS:
pip install -r requirements.txt export APIFY_TOKEN="your-token" python run_workflow.py "led strip lights" "sunset lamp" "galaxy projector" --advertisers govee.com blisslights.com Windows PowerShell:
pip install -r requirements.txt $env:APIFY_TOKEN = "your-token" python run_workflow.py "led strip lights" "sunset lamp" "galaxy projector" --advertisers govee.com blisslights.com The Actors: AliExpress Product Research Scraper · Google Trends Scraper · Google Ads Transparency Scraper.
If something in the numbers looks wrong, tell me in the comments. I'd rather fix the method than defend it.
