"What should I sell?" is the question every new seller gets stuck on. Lists of "trending products" don't help much: by the time a product is on everyone's list, hundreds of sellers are already fighting over it.
A better approach is a repeatable process that turns ideas into evidence. Here are the seven steps we recommend, whether you sell on Amazon, Shopify or TikTok Shop.
Step 1: Set your constraints first
Before you look at a single product, write down:
- Market: one country to start (for example the US, UK or Germany). Fees, VAT and demand differ by country.
- Channel: Amazon FBA, your own Shopify store, or TikTok Shop. Each has different fees and different buyers.
- Budget: what you can spend on the first order, samples and launch ads together.
- Rules you won't break: for example nothing fragile, nothing electrical, nothing that needs certification you can't get.
These constraints remove most bad ideas before they cost you time.
Step 2: Generate ideas from real buying behaviour
Good idea sources show what people already buy:
- Amazon best-seller and "movers and shakers" lists in your category.
- Google Shopping results for broad terms in your niche.
- Problems you see in reviews of popular products ("I wish it came in a bigger size").
- Things you or people around you buy repeatedly.
Aim for 20–30 ideas. You will drop most of them.
Step 3: Check demand with data, not opinions
For each idea, look for at least two independent demand signals:
- Search demand: monthly search volume for the main keyword, and its 12-month trend. Remember that Google Trends shows relative interest on a 0–100 scale, not absolute searches.
- Sales evidence: for Amazon, a product's Best Seller Rank (BSR) history shows whether it sells steadily or only spikes. A lower rank means more sales within that category.
- Reviews over time: many recent reviews suggest ongoing sales.
One signal can mislead. Two that agree are much more convincing.
Step 4: Size up the competition
Demand only helps if you can win some of it. For the top listings, note:
- How many reviews the leaders have. A wall of listings with thousands of reviews is a moat.
- How many sellers offer the same item.
- Whether listings look professional or weak. Weak images and confusing titles are an opening.
- How much brands spend on ads (search the Meta Ad Library and look at sponsored results).
Step 5: Do the margin math before you fall in love
Work out the profit per unit at a realistic price:
Selling price − VAT − product cost − shipping − import duty − marketplace fees − fulfilment − ad cost per sale = profit per unit.
If the margin is thin before ads, it disappears after ads. See our step-by-step guide to calculating Amazon FBA profit in the US, UK and Germany.
Step 6: Look for risks that kill launches
- Seasonality: a product that sells only in December locks up cash the rest of the year.
- Price wars: a falling price history means sellers are competing on price.
- Returns and breakage: fragile or sized items come back more often.
- Compliance: toys, electronics, cosmetics and food contact items often need testing or certification.
Step 7: Shortlist, validate, then test
Rank the survivors on demand, trend, competition, margin and risk, and pick the top one to three. Then validate the best one in depth and run a small ad test before you order inventory (see ad creatives to test before you buy inventory).
A simple scoring sheet
| Score (0–100) | Higher is better? | Based on |
|---|---|---|
| Demand | Yes | Search volume, sales rank, review velocity |
| Trend | Yes | 12-month search and sales direction |
| Competition | No | Review moats, seller count, ad spend |
| Margin | Yes | Profit after all fees at a realistic price |
| Risk | No | Seasonality, price drops, returns, compliance |
Leave a score empty when you don't have the data. A guessed number is worse than an honest gap.
Doing this faster
This process works, but checking 30 ideas by hand takes days. Ecomtoro runs steps 3 to 6 for you: it ranks product ideas using Amazon sales history, search demand and competition, calculates margins after fees, VAT and duty, and shows the source behind every number. You can try it free.
Frequently asked questions
How many product ideas should I research before choosing one?
Start with 20–30 ideas and filter them with your constraints and demand data. Most will drop out quickly; you only need one to three strong candidates to validate in depth.
Is a high search volume enough to pick a product?
No. High search volume often means strong competition. Combine demand with competition, margin and risk before deciding, and confirm demand with a second signal such as sales-rank history.
Should I sell on Amazon, Shopify or TikTok Shop first?
Amazon brings existing buyer traffic but charges referral and fulfilment fees; Shopify gives you more control but you must drive all traffic; TikTok Shop suits products that demonstrate well on video. Pick one, and calculate the margin for that channel.
