Start with a simple, non-fragile, non-electronic product you can differentiate from competitor complaints, priced high enough to absorb advertising. Check that it can be sold through more than one traffic channel. If the only way to sell it is Amazon search, your entire launch depends on winning one auction against people who got there first.
The short version
- Simple beats clever for a first product. Fewer parts, fewer failure modes, fewer returns.
- Differentiate from complaints, never from invention. The negative reviews on page one are your product brief.
- Price has to carry advertising. A low ticket item leaves nothing to acquire a customer with.
- Avoid fragile, electronic, regulated and sized-to-fit. Each adds a return rate you have not learned to manage yet.
- Ask how you would sell it if search disappeared. A product with only one route to a buyer is a fragile business.
Work this checklist in order. Every item has a defined meaning of done, and a first product should clear all eight before you request a single quote from a factory.
The eight item checklist
- The category is big enough to be worth entering. Done properly means you have written down category revenue, growth trajectory and how concentrated the top sellers are, before you have any emotional attachment to a product.
- The product is physically simple. Done properly means no motors, no batteries, no firmware, no glass, and nothing that needs sizing charts. Every one of those raises return rate, and return rate quietly eats the margin you were counting on.
- There is a rating gap you can attack. Done properly means the leaders on page one sit somewhere short of excellent, and the reasons appear repeatedly in their one and two star reviews.
- The differentiation is written as a sentence. Done properly means you can say what you are changing and which specific complaint it answers. We build differentiation from competitor negative reviews and the rating gap, never from inventing a feature nobody asked for. If your sentence starts with "what if it also", start again.
- The price supports acquisition. Done properly means you have subtracted product cost, Amazon fees, freight and an honest advertising allowance, and there is still margin left. Cheap products fail here more often than they fail on demand.
- It can travel down more than one traffic channel. There are five: organic search, paid advertising, promotions, influencer and creator content, and off-channel traffic. Most sellers run two and then wonder why growth stops. Done properly means you can name at least three that suit this specific product, and creator content in particular needs a product that is visually demonstrable.
- Supply is not concentrated in one factory. Done properly means you have identified more than one credible manufacturer, because a single-source product hands your pricing and your timeline to someone else.
- You can afford to test it without flinching. Done properly means the first phase, about 200 units and $5,000 to $10,000, is money you can lose. We test up to four products at once at that scale precisely because most of the value is in comparison.
The categories I steer first-time sellers away from
Supplements and anything ingestible, because of compliance and liability. Apparel and footwear, because sizing drives returns. Consumer electronics, because warranty and support obligations arrive later and never leave. Anything requiring certification you do not already understand. None of these are unwinnable. They are bad places to learn, because the mistakes are expensive and slow to surface.
What product research tools will not tell you
Software will show you revenue estimates, review counts and price history. It will not show you the two things that decide a private label launch: whether the incumbents are actually vulnerable, and what it costs to acquire a customer in that category. Both require reading, not filtering. The research sheet we fill in per opportunity is long, and the entries that carry the most weight are qualitative, like why customers are unhappy and whether that unhappiness is fixable at your target cost.
The second thing tools hide is competitive intent. A category can look soft and be defended by an operator who will drop price the week you appear. You find that by reading their pricing history and their review response behavior, not by sorting a spreadsheet by opportunity score.
Related answers
- How to validate a product before Amazon launch
- What tools to use for Amazon product research
- How to source from manufacturers safely and verify quality
- Global launch vs single marketplace on Amazon which first
- Amazon launch services: the complete guide
Send us a shortlist and we will tell you which ones we would refuse to launch, at Flapen.
