How to Do Product Research for Amazon FBA & Dropshipping

Product research is how you decide what to sell before you put money behind it. Get it right and everything after it gets easier, from finding a supplier to running your first ad. Get it wrong and no amount of marketing will save a product people don't want, or one you can't sell at a profit.

The approach holds up across selling models. An Amazon FBA private label and a dropshipping test both come down to the same four questions. Is there real demand? Can you compete? Do the numbers actually work? And is there a gap you can move into? What changes between the two models is where you get the data and which costs do the most damage to your margin. This guide walks through all four, with a deeper guide linked for each step.

You're not looking for a winner yet. You want a shortlist of products worth the ten minutes of scrutiny the next steps take. Cast a wide net first, then let the data cut it down.

On Amazon

• Read the category Best Sellers and Movers & Shakers lists, and favour products with steady demand over one-off spikes.
• Pick a keyword you understand and study the listings ranking for it. Weak or dated listings on page one are an opening.
• Look next to products you already know. Accessories, bundles, and adjacent niches carry less risk than a category you've never touched.

In dropshipping

• Check what already sells on AliExpress and Temu. The order counts on a supplier listing are a public demand signal you can read before spending anything.
• Watch short-form video on TikTok and Reels for products with an obvious hook, either a wow factor or a clear problem they solve.
• Track rising search interest so you catch a trend while it climbs, not once it peaks and every store is running it.

Aim for ten to twenty candidates. Most will die in the next two steps, which is exactly what they're for. If you want more places to look, the guide to finding winning products below goes through each source in detail.

Demand is the question that kills most products. Are enough people actually buying this, week after week, right now? Likes and search volume look like demand, but they aren't. Sales are. So get as close to real sales figures as each platform lets you.

On Amazon that means reading Best Sellers Rank as a stand-in for sales speed, checking it against the "bought in past month" figure where Amazon shows one, and looking at a few weeks of history instead of a single day. In dropshipping it means the order count on the supplier listing, the engagement on ads already running the product, and which way search interest is heading.

The demand guide below covers the full method, including how to turn a BSR into a rough unit estimate and how to read dropshipping signals without talking yourself into a bad idea.

A product can have strong demand and still be a poor bet if the field is impossible to break into. Demand tells you the opportunity is real. Competition tells you whether you can actually reach it.

• On Amazon, count the sellers on the listing and look at the review counts on the top results. A first page where every listing has thousands of reviews is a wall you probably can't climb. Watch for one brand that dominates, or for Amazon selling the product itself.
• In dropshipping, judge how saturated the product is. How many stores already run it, how stale the ad creative looks, and whether you can stand out on the offer, the angle, or a bundle instead of racing everyone else to the lowest price.

What you want is a gap: proven demand where the sellers already there are weak, slow, or beatable on something you can control. The competition guide below breaks down what to check on each platform.

This is where a lot of exciting products quietly fall apart. Strong demand and light competition still add up to nothing if there's no profit left after every fee, ad dollar, and return.

The cost stack looks very different by model. Amazon FBA turns on referral fees, fulfilment fees, and the landed cost of your goods. Dropshipping turns on ad spend, because getting the click is usually your biggest cost and the one most likely to eat a margin that looked healthy on paper. Run the real numbers for your model before you order stock or launch a campaign. The margins guide below walks through both.

Product research is a filter, not a hunt for certainty. Run every candidate through the same four questions and you'll drop the losers early, backing only the products that clear all of them. Do that consistently and you've handled most of the job.