The problem
Why most "winning products" lose money
Every week thousands of new stores launch with a product someone called a winner in a video, a Facebook group or a spy tool. Most of those tests end the same way: a few hundred dollars of ad spend, a handful of sales, and a product that never had the economics to work.
The pattern behind the losses is consistent. A product gets picked because it looks popular, but nobody checks three things that actually decide the outcome:
- Is the demand proven, or is it one viral video riding on novelty?
- Does the margin survive advertising? A $20 product with a $10 contribution cannot pay a $14 cost per customer.
- Is it still differentiated, or are fifty stores already running the same creative to the same audience?
Winning product research is the discipline of answering those questions before you buy traffic. Done well, it does not guarantee a hit. It removes the products that were never going to work, so your test budget goes to the few that can.
The framework
The 9 signals of a winning product
Every serious method for finding winning products ends up checking some version of the same nine signals. ProductScore rates each one from 1 to 4 and combines them into a 0–100 Opportunity Score. The percentages below are the weights used in the General research mode; the TikTok, Meta, Google Shopping, niche-store and high-margin modes shift them toward what each channel rewards.
Demand 18%
Lifetime orders on AliExpress are the cleanest public proof that people buy. Under 100 orders is cold, 500+ is proven, 2,000+ is the top tier.
Problem solving 18%
Products that fix a real, felt annoyance (pain, mess, clutter, wasted time) convert better and need less persuasion than novelties.
Video potential 14%
Can the value be shown, not told, in the first three seconds? A before-and-after or a satisfying transformation is worth more than any copy.
Impulse purchase 14%
Low enough price and strong enough emotion to buy without research. In practice the sweet spot sits around $19.99 to $39.99.
Differentiation 10%
Not sold in every supermarket or on Amazon's first page. A unique angle, design or bundle lets you hold your price.
Dropshipping suitability 10%
Light, compact, durable and simple to ship. Fragile, oversized, perishable or battery-heavy items cause refunds and chargebacks.
Ad creative potential 6%
Several distinct hooks and angles to test, so one fatigued ad does not end the product's life.
Competition 5%
How crowded the listing pool is. Many near-identical listings mean a race to the bottom on price.
Product risk 5%
Regulatory, IP, safety and platform-policy exposure. Medical claims, branded look-alikes and restricted items get ad accounts banned.
Notice what carries the weight: demand and problem solving together are more than a third of the score. A product that is merely interesting rarely scores well. A product that solves something people already pay to fix usually does. The full rubric for each dimension, and how each research mode shifts the weights, is in the ProductScore score guide.
Step by step
How to find winning products on AliExpress
Here is the process that works whether you research by hand or with a tool. Budget about an hour per niche if you do it manually.
Start from a niche, not a single product
Pick a niche where people spend repeatedly and feel a problem: pet care, home organization, sleep, posture, car accessories, kitchen gadgets, beauty tools. A niche gives you related products to test and a store that makes sense to buyers.
Pull the listings and sort by orders
Search the niche on AliExpress and collect the top listings by order count. Note price, orders, rating, shipping options and how many sellers offer the same item.
Cut everything without proven demand
Remove listings with fewer than 100 lifetime orders unless you have a strong reason to be early. Prefer 500+ orders. Treat 2,000+ as proven, but check whether that demand is now spread across dozens of copycat stores.
Run the margin check
Multiply the landed cost (product plus shipping) by 2.5 to get a target retail price. If that price looks wrong for the product, or lands far outside the $19.99–$39.99 impulse band, the product will struggle on paid social.
Calculate your break-even CAC
Subtract product cost, shipping, payment fees and a returns allowance from the retail price. What is left is the most you can pay for one customer. Compare it with typical CAC for your ad channel (see the table below). Use the free break-even CAC calculator if you do not want to build a spreadsheet.
Score the qualitative signals
For each survivor, rate problem solving, video potential, impulse appeal, differentiation and ad-creative potential. Be honest: would you stop scrolling for this? Can you name three different hooks?
Screen out the risks
Drop anything fragile, oversized, regulated, trademark-adjacent or likely to break in transit. One problem product can cost more in refunds and account bans than a dozen good ones earn.
Shortlist and test
Keep the top three to five products and test them with small, equal budgets. The research job ends here; the creative job starts.
Skip the spreadsheetSteps 2 to 7 are what ProductScore runs for you: live AliExpress listings, 9-dimension AI scoring, safety gates and break-even CAC in one results table.
Score a niche freePrint it, save it
The winning product checklist
Use this before you order samples or build a product page. A real winner ticks at least ten of the fourteen boxes, and never fails the last three.
- 500+ lifetime orders on AliExpress (100 is the absolute floor)
- Solves a problem the buyer already feels, or delivers an obvious "wow"
- The benefit is visible in the first three seconds of a video
- Retail price of roughly $19.99–$39.99 after a 2.5x markup
- Contribution after costs and fees is higher than your channel's typical CAC
- Not on every Amazon first page or in local supermarkets
- You can write at least three distinct ad hooks for it
- Room to bundle, upsell or offer a variant
- Reliable supplier: good rating, tracked shipping, consistent photos
- Similar listings make up a small share of the niche, not a wall of clones
- Broad audience, or a narrow audience that is easy to target
- Light, compact and durable enough to ship without breakage
- No medical, safety or trademark claims needed to sell it
- Not restricted by TikTok, Meta or Google ad policies
The last three items are pass/fail. In ProductScore they are enforced as safety gates that cap a product's score no matter how well it does elsewhere.
Unit economics
The profit math that decides whether a product wins
The fastest way to kill a "winning product" is to run its numbers. Take two products, both priced with the same 2.5x markup, 2.9% + $0.30 payment fees and a 3% returns allowance:
| Per order | $8 product | $14 product |
|---|---|---|
| Retail price (2.5x) | $20.00 | $35.00 |
| Product + shipping | −$8.00 | −$14.00 |
| Payment fees | −$0.88 | −$1.31 |
| Returns allowance (3%) | −$0.60 | −$1.05 |
| Contribution before ads | $10.52 | $18.64 |
| Break-even ROAS | 1.90 | 1.88 |
| Profit at TikTok CAC ($9.50) | +$1.02 | +$9.14 |
| Profit at Meta CAC ($14.00) | −$3.48 | +$4.64 |
| Profit at Google Shopping CAC ($17.50) | −$6.98 | +$1.14 |
Same markup, very different businesses. The $8 product only works on the cheapest channel and has almost no room for a bad day. The $14 product survives on all three. This is why the price you can charge matters more than the supplier price you pay, and why a 2.5x markup is a floor, not a target.
Rule of thumb: if your contribution before ads is not comfortably above your channel's typical CAC, it is not a winning product for that channel, no matter how many orders it has. Check yours with the free dropshipping profit calculator or break-even ROAS calculator.
Want the long version? How to price dropshipping products and break-even CAC and ROAS explained walk through every line of this table.
Channel fit
TikTok winning products vs. Meta vs. Google Shopping
A product can be a winner on one channel and a loser on another. The traffic source decides what the buyer needs to see and how much each buyer costs.
TikTok Ads and TikTok Shop
TikTok rewards instant visual payoff: a transformation, a satisfying result, a "how did I live without this" moment. Impulse prices work best and baseline CAC is the lowest of the three. Weak points: creative fatigues fast and copycats arrive within weeks. Read TikTok Shop vs. Meta ads product selection for the trade-offs.
Meta (Facebook and Instagram)
Meta tolerates higher prices and products that need a little explanation, especially for audiences over 30. You pay more per customer, so you need more margin. Problem-solving products with a clear demo do well here.
Google Shopping
Google Shopping captures people already searching for a product by name. That favours products with existing search demand and clear specs over pure novelty, and it is the most expensive of the three per customer. See Google Shopping dropshipping products.
Baseline CAC by market
These are the estimated customer acquisition costs ProductScore uses as starting points. They are planning baselines, not quotes: your real CAC depends on creative, offer and audience.
| Market | TikTok Ads | Meta (FB/IG) | Google Shopping |
|---|---|---|---|
| United States | $9.50 | $14.00 | $17.50 |
| United Kingdom | $8.55 | $12.60 | $15.75 |
| Germany | $7.60 | $11.20 | $14.00 |
| GCC / Arab Gulf | $6.65 | $9.80 | $12.25 |
| Morocco | $3.32 | $4.90 | $6.12 |
Full methodology and more markets on the CAC benchmarks page. Background reading: dropshipping CAC by country.
What to avoid
Red flags: saturated and risky products
Spotting a loser early is worth as much as spotting a winner. These are the patterns that most often turn a promising product into a refund queue.
Cold demand
Fewer than 100 lifetime orders. Sometimes that is an early opportunity, usually it means nobody wants it. ProductScore caps cold-demand products at a score of 40.
Saturation
The same product from dozens of sellers, identical ads across many stores, falling prices and comments like "I saw this for half the price". ProductScore measures the share of similar listings in each search: under 5% is open water, above 30% is a crowded pool. More in how to tell if a dropshipping product is saturated and the product saturation definition.
Dropshipping blockers
Fragile glass, liquids, large or heavy items, sizing-heavy apparel and anything needing assembly. Long shipping plus breakage equals chargebacks. These products are capped at 55.
Policy and legal risk
Health and medical claims, weight-loss promises, weapons-adjacent items, brand look-alikes and children's products without certification. They get ads rejected and accounts restricted. High-risk products are capped at 50.
Novelty without a problem
Gets views, not purchases. If the comments say "cool" but nobody asks "where can I buy it", the product is entertainment, not a winner.
Before you spend on ads, run the checks in how to validate a dropshipping product before ads.
Sources
Where to look for winning products in 2026
Product ideas are cheap; validation is expensive. These sources are good for finding candidates. None of them tells you whether a product is profitable for you, which is why each idea still needs the checks above.
- AliExpress order counts and new arrivals. The most direct demand signal you can see for free, and the source ProductScore reads live.
- TikTok Creative Center and the TikTok search bar. Shows what is getting engagement right now, and which hooks are being used.
- Meta Ad Library. Ads that have been running for weeks are usually profitable for someone. Count how many stores run the same product.
- Amazon Best Sellers and Movers & Shakers. Good for spotting demand, but a product on Amazon's first page is hard to differentiate.
- Google Trends. Separates steady demand from a seasonal spike or a fad that already peaked.
- Reddit and niche forums. People describing problems in their own words is the best ad copy you will ever find.
Scored niche pages
ProductScore publishes scored niche pages built from its own research runs, with products ranked by Opportunity Score. A few to start with:
- Night light winning productsScored AliExpress listings in the night-light niche
- Bluetooth speaker productsDemand, margin and risk for portable audio
- Smart home device productsGadgets scored for dropshipping suitability
- Wall art productsHome decor listings ranked by opportunity
Browse every niche on the winning products directory or see what is moving this month on trending products.
The tool
ProductScore: the AI winning product finder
ProductScore turns the process on this page into one search. You type a niche, pick a research mode, and get a ranked table of live AliExpress listings with a score, a verdict and the numbers behind both.
Search a niche or paste a listing
ProductScore pulls current AliExpress listings with prices, order counts and ratings. Choose one of 6 research modes: General, TikTok Ads, Meta, Niche Store, High Margin or Google Shopping.
Every product is scored on 9 dimensions
Demand and competition come from the listing data. Problem solving, video potential, impulse appeal, differentiation, dropshipping suitability, ad-creative potential and risk come from a calibrated AI rubric. Safety gates then cap anything risky, unshippable or unproven.
See the economics, not just the score
Each product shows retail at 2.5x, contribution, break-even CAC and break-even ROAS against your channel, plus a plain verdict. Save, compare and shortlist the best ones.
What the score means
Strong on most signals. Worth testing first.
Shortlist material. Check the weak dimensions.
Only with a specific angle or audience.
Or any product that fails a safety gate or the margin check.
Manual research vs. ProductScore
| Task | By hand | With ProductScore |
|---|---|---|
| Collect listings and orders | Copy from AliExpress into a sheet | Pulled live per search |
| Rate 9 signals | Your judgement, inconsistent between days | Same AI rubric for every product |
| Margin and break-even CAC | Spreadsheet formulas per product | Calculated for each listing and channel |
| Risk screening | Easy to skip when excited | Safety gates applied automatically |
| Saturation check | Manual count of similar listings | Share of similar listings measured per batch |
| Time per niche | About an hour | A few minutes |
Curious how the engine works under the hood? How ProductScore works explains the data pipeline and scoring step by step.
Pricing
Start with the free plan
The free plan is available today and includes everything you need to try the method on this page. Paid plans with higher limits are on the way.
Free
$0 forever
- 3 researches / month
- Up to 20 products each
- Opportunity Score + AI breakdown
- Safety gates, all 6 modes
- Break-even CAC
Starter
$19 / month
- 50 researches / month
- Up to 50 products each
- CSV exports
Pro Creator
$49 / month
- 200 researches / month
- Up to 100 products each
- Custom scoring formulas
Agency
$119 / month
- 500 researches / month
- Up to 200 products each
- REST API access
Details and the latest limits on the ProductScore pricing page.
Find your first winning product todayCreate a free account and score a niche in a few minutes. No credit card.
Create free accountKeep learning
Free tools and guides for product research
- Dropshipping profit calculatorPer-order profit after fees, returns and ads
- Break-even CAC calculatorThe most you can pay to acquire a customer
- Break-even ROAS calculatorThe return on ad spend you need to not lose money
- CAC benchmarks by marketBaseline costs for TikTok, Meta and Google
- How to spot winning AliExpress productsThe 2026 deep dive on AliExpress research
- The 3x margin ruleWhy markup rules exist, and where they break
- Dropshipping glossaryOpportunity Score, landed cost, contribution margin and more
- ProductScore blogResearch methods, channel guides and economics
FAQ
Winning products: frequently asked questions
What is a winning product in dropshipping?
A winning product is one that can be sold profitably through paid ads at scale: it has proven demand, a clear problem or hook that works on video, enough margin to cover the cost of acquiring a customer, and low risk of shipping, quality or policy problems. Popularity alone does not make a product a winner; profitable demand does.
How do I find winning products on AliExpress?
Search a niche, keep listings with proven order volume (500+ lifetime orders is a good bar), check that the retail price you can charge is at least 2.5x the landed cost, confirm the product shows its value in a short video, estimate your break-even CAC, and screen out fragile, regulated or oversized items. ProductScore automates these steps and returns a 0–100 Opportunity Score for each listing.
How many orders should a winning product have on AliExpress?
As a rule of thumb, under 100 lifetime orders is cold demand, 500 or more is proven demand, and 2,000 or more is the top demand tier. ProductScore caps the score of cold-demand listings at 40 so an unproven product cannot look like a winner.
What profit margin does a winning product need?
Your contribution per order, after product cost, shipping, payment fees and returns, must be higher than what you pay in ads to win a customer. A 2.5x markup and a retail price between $19.99 and $39.99 is a practical starting point. For example, a product that costs $14 and sells for $35 leaves about $18.64 before ads, which covers a typical Meta CAC; an $8 product sold at $20 leaves only $10.52 and loses money at a $14 CAC.
How can I tell if a winning product is already saturated?
Look for many near-identical listings from different sellers, the same ad creative running from dozens of stores, falling prices and comment sections full of people saying they saw it cheaper elsewhere. ProductScore measures the share of similar listings in each search batch and lowers the competition score when that share rises above 5%, 15% and 30%.
Are TikTok winning products different from Facebook winning products?
Partly. TikTok rewards products with an instant visual payoff and impulse prices, and its baseline CAC is lower. Meta tolerates slightly higher prices and more explanation but costs more per customer. Google Shopping favours products people already search for by name. That is why ProductScore has separate research modes for TikTok Ads, Meta, Google Shopping, niche stores and high-margin products.
Is there a free winning product finder?
Yes. ProductScore has a free plan with 3 researches per month and up to 20 products per research, including the Opportunity Score, the AI rubric breakdown, safety-gate screening, all 6 research modes and break-even CAC estimates. No credit card is required.
Does a high score guarantee that a product will sell?
No. A score is a filter that tells you which products deserve an ad test and which ones to skip. Creative, offer, store, timing and execution still decide the result. The value is in testing fewer, better products and avoiding the obvious losers.
Who publishes winningproduct.si?
winningproduct.si is a free guide published by the team behind ProductScore (productscore.si), an AI product research tool for AliExpress dropshipping. Questions? Email [email protected].