To find winning products, start from evidence, not from lists. Look for products that several stores keep advertising, check that people want them outside of ads too, and make sure the price leaves room for ads, shipping and fees. Then run a small test with a few different ad angles and keep only what sells for you.
A "winning product" is simply a product that sells at a profit for your store, with ads that keep working when you spend more. Nobody can give you one. What you can do is cut the number of tests that start from a guess by checking the same few things every time. This guide walks through those checks, with a real example from our data.
What makes a product worth selling?
A product has to pass four filters. If one fails, the product usually fails with it.
| Filter | Question | Evidence to look for | Warning sign |
|---|---|---|---|
| Demand | Do people already buy this kind of product? | Several stores advertising it for weeks; interest outside ads | One viral video, nothing else |
| Room | Can you enter with your own angle? | Ads use different hooks and offers | Every store runs the same video and discount |
| Margin | Is there money left to pay for ads? | Selling price far above landed cost | You need a perfect week to make a profit |
| Execution | Can you show it, ship it and support it? | Easy to demo; reliable supplier; few sizes or variants | Fragile, confusing or slow to arrive |
The order matters. Many sellers pick a product they like first and then look for data that agrees. Start with the evidence, then ask if the product is something you can sell well.
Where do good product ideas come from?
Ideas come from three kinds of places. Each shows a different kind of behaviour.
| Source | Examples | What it shows | Watch out for |
|---|---|---|---|
| Paid ads | Meta Ad Library, ad research tools | What stores keep paying to promote | Long-running ads can have tiny budgets |
| Purchases | Amazon Best Sellers and Movers & Shakers, Etsy, eBay | What people buy | Marketplace demand may not move to your store |
| Attention | TikTok, Instagram, Reddit, Facebook groups | Problems, questions and words buyers use | Views and likes are not orders |
Amazon's Movers & Shakers page lists the biggest gainers in sales rank over the past 24 hours, so it is useful for spotting sudden moves, not for judging lasting demand. Communities are best for the buyer's language: our guide to product research on Reddit shows how to search them.
A simple rule: an idea that appears in only one group stays an idea. When it appears in two groups, for example ads from several stores plus steady marketplace sales, it becomes a candidate.
How to find winning products, step by step
- Start from a problem, not a product. "Sore back after work" gives you more products and angles than "posture corrector". Write down three problems in a niche you understand.
- Search the ads. In Meta's Ad Library, search the problem words and the product words, one country at a time. Each card shows the ad's start date, and Meta's help page explains the filters. Our guide on finding products on Facebook covers this step in detail.
- Count advertisers, not ads. One brand can run hundreds of ads. A product sold by many different stores is stronger evidence.
- Read the ads that last. Note the hook, the format (demo, UGC, before and after), the offer and the start date of the ads that have run for weeks. These are the ads that work for someone, and they show you the angles buyers respond to.
- Open the store. Check the price, the bundle, the guarantee, the shipping promise and how the page answers doubts. A weak page with long-running ads can mean the demand is strong enough for a better store to win.
- Check demand outside ads. Look at Google Trends over 12 months. Google scales its numbers from 0 to 100 against all searches, and very small terms show as 0 (Google Trends Help), so compare the shape of the line, not the number.
- Do the margin math. See the next section. Drop the product if ads can't fit in the margin.
- Check suppliers. Order samples from two suppliers if you can. Check quality, packaging, variants and real delivery times. If you source on AliExpress, our guide on how to contact a seller on AliExpress helps you ask the right questions.
- Run a small test. One product, one clear offer, three to five different angles, and rules for stopping that you write down before you start.
A worked example: red light therapy (October 2026)
Here is how steps 2 to 5 look with real data.
- On 26 October 2026, the keyword "red light therapy" had 3,052 active Meta ads from 519 advertisers, selling 307 products. That is demand evidence: hundreds of stores pay to reach these buyers.
- Sorted by estimated revenue, two of the leading products were Omnilux Contour Face ($610K–$1.5M a month, 260K store visits in September 2026) and the Megelin Red and Infrared Light Therapy Bag ($550K–$1.3M, 70K store visits). Revenue is our estimate, shown as a range.
- 96% of Omnilux's ads and 93% of Megelin's targeted the US.
What does that tell a new store? The category sells, but the leaders are established brands focused on one market. That gives you hypotheses to check, not answers: another country where few stores run these ads yet, a narrower buyer, or an accessory around the main product. Each one goes back through the same filters. Health products also face stricter ad rules: Meta lists health and wellness among its restricted categories (Meta Advertising Standards), so check what claims you can make before you build anything.
How do you check the margin before you test?
Use break-even ROAS: the return on ad spend at which you neither win nor lose money.
Break-even ROAS = selling price ÷ (selling price − product cost − shipping − fees − returns allowance)
Example with made-up numbers:
| Line | Amount |
|---|---|
| Selling price | $40.00 |
| Product cost | $10.00 |
| Shipping to the customer | $6.00 |
| Payment fees | $1.50 |
| Returns and replacements allowance | $1.50 |
| Left to pay for ads and profit | $21.00 |
| Break-even ROAS | 40 ÷ 21 = about 1.9 |
So every $1 of ads must bring about $1.90 in sales just to break even. If the ads in your niche can't realistically reach that, the product fails the margin filter, however good the ads look. Our break-even ROAS calculator does this math for you.
How do you read your first test?
Read the results in order, from the ad to the purchase. Each stage points to a different fix.
| What you see | Likely problem | What to try |
|---|---|---|
| Few clicks on every angle | The hook or the product itself | New hooks; if nothing works, drop it |
| Clicks, but few add-to-carts | The page or the offer | Clearer promise, proof and price above the fold |
| Add-to-carts, but few purchases | Price, shipping or trust at checkout | Show delivery times, returns and total price early |
| Sales, but no profit | Margin or order value | Bundles, a higher price, or a cheaper supplier |
Then put every test in one of three groups:
- Drop when no angle gets attention.
- Fix when the ad works but the page or offer leaks buyers.
- Scale slowly when several angles sell and the margin holds. Add new creatives as you go, because one ad rarely carries a product for long.
Which products usually fail the checks?
- Products that only look good in a perfectly edited video and disappoint in real use.
- Products with no angle except "cheaper than Amazon".
- Big, heavy or fragile items where shipping eats the margin.
- Products with many sizes or fits, which bring returns.
- Lookalikes of branded or patented products, which bring rejected ads and legal risk.
- Products that solve a problem people don't care enough about to pay for.
Keep a research log
Write one line per product you looked at: where you found it, why it looked good, which filter it failed or passed, what happened in the test and what you would do differently. After a few months the log shows you patterns that no tool can: which angles work for your audience, which suppliers deliver, which kind of product your store sells best.
For products that are growing right now, see our guide to trending dropshipping products.
How to do this with SearchTheTrend
SearchTheTrend shows the Meta ads that keep running, the products behind them and the stores that sell them.
- In Products, filter by niche, country, estimated revenue or active ads, and sort by what is growing.
- Open a product to see its estimated monthly revenue range, monthly visits, market by country, weekly store price and every ad that sells it.
- From each ad, check days running, format and Ad Rank to find the angles brands keep paying for. For ads shown in the EU or UK you also see estimated spend in Europe.
- Open the store to see its other best sellers, traffic and tech.
It covers Meta ads only, not TikTok. The free plan needs no card.
FAQ
What is a winning product?
A product that sells at a profit for your store, with ads that keep working as you spend more. No list or tool can promise one: you collect evidence of demand, check the margin, and let a small test decide.
Can I do product research for free?
Yes. Meta's Ad Library, Google Trends, Amazon's best-seller and Movers & Shakers pages and online communities cost nothing. Paid research tools save time by sorting products, keeping ad history and linking each ad to its store.
How do I know if a product is saturated?
Look at the ads, not the number of sellers. If every store runs the same video, the same promise and the same discount, the market is crowded. If the ads use different angles and new advertisers keep arriving, there is still room.
How much margin does a product need?
Enough to pay for the ads that sell it. Work out your break-even ROAS (selling price divided by what is left after product, shipping and fees) and compare it with the return your ads can realistically reach.
Do these steps work for products I find on TikTok?
Yes, the checks are the same. Just note that SearchTheTrend covers Meta ads only (Facebook and Instagram): a TikTok product shows up there only if stores also advertise it on Meta.
