To find products that sell, look for the same opportunity in several places: steady search interest, real sales, repeated customer complaints and room for a better offer. Start with Amazon best sellers and Google Trends. Then read reviews, Reddit threads and social comments to learn what buyers still want. Compare your strongest ideas with the scorecard below. Test buyer interest before paying for design, tooling or inventory.
A promising product usually has five signs:
People already search for and buy comparable products.
The same unmet need appears in reviews and customer conversations.
You can offer a difference customers will understand.
Expected costs leave room for profit at a realistic selling price.
A manufacturer can produce it at a workable quantity, cost and timeline.
How to Find Products That Sell
Finding a product with sales doesn’t mean you’ve found a good product to build. You need enough evidence to explain who buys it, why they buy it and where current options fall short.
One of the first things we look for is agreement between different kinds of evidence. Search data can show interest. Marketplace sales can show buying behavior. Reviews can tell you what customers dislike. Conversations can reveal problems that keyword tools will never show you.
The strongest opportunities usually appear in more than one place.
For example, you may find a category with steady Amazon sales. That gets your attention. Then you notice the same complaint across dozens of product reviews. Reddit users describe the same frustration and explain the workaround they use instead. That starts to look like a product problem worth studying.
Now compare the opportunity with the business behind it. Estimate the selling price, landed cost, marketplace fees, returns and customer acquisition expense. Find out what minimum order quantities suppliers expect. Check whether testing or certification requirements could change the economics.
You don’t need final factory quotes while you’re still comparing ideas. You do need enough information to spot an idea that makes little financial or production sense.
If you’re still learning the broader research process, Gembah’s product research guide covers that subject in more depth. This article stays focused on finding product opportunities and deciding which ones deserve more attention.
Eight Ways to Find Product Opportunities
No single research method tells you enough. Amazon can show what sells today while customer conversations reveal what buyers wish existed. Use several methods, then look for patterns that repeat.
1. Study Amazon Best Sellers for Demand Patterns
Amazon gives you a direct look at categories where people already spend money. Start with Best Sellers, Movers & Shakers and relevant search results. Look at price ranges, review counts, product variations and how many brands appear repeatedly.
Don’t hunt for one hot listing to copy. Study the category.
Pay attention to products that keep selling without one brand controlling every visible result. Then read reviews. A category becomes more interesting when buyers keep spending money while complaining about the same feature, material, fit or usability problem.
Amazon research tools can speed up this work. The right choice depends on how much data you need.
| Tool | Useful for | What to examine |
|---|---|---|
| Helium 10 | Private-label and multi-product research | Estimated sales, keywords, competing ASINs and category patterns |
| Jungle Scout | Product discovery and Amazon demand research | Sales estimates, competition and supplier information |
| SmartScout | Brand and category research | Market share, brand concentration and category structure |
| Viral Launch | Product and keyword research | Competitor listings, keyword demand and category activity |
Treat every sales estimate as an estimate. You’re looking for enough evidence to compare opportunities, not a perfect prediction of future revenue.
2. Use Google Trends to Check Staying Power
Google Trends can help you tell the difference between sustained interest, seasonality and a short-lived spike.
Start with a five-year view. That gives you enough history to see whether the category is growing, flat or fading. Then switch to the past 12 months to see seasonal patterns.
Compare the broad category with specific product variations. A large category can be flat while one use case or feature is gaining interest.
Related Queries can also reveal how customer language is changing. Searches involving complaints, alternatives, specific audiences or new uses may point toward a narrower product opportunity.
Don’t reject a seasonal product just because demand rises and falls. Seasonal demand can work well when you understand when buyers shop and can plan inventory around it.
3. Read Reddit and Niche Forums for Unsolved Problems
Search data tells you what people type. Communities often tell you why they’re frustrated.
Look through subreddits, hobby forums and professional communities related to the category. Pay close attention to recommendation threads, complaints and posts where people describe their own fixes.
Phrases such as “I wish this had,” “does anyone make” or “I had to modify mine” can point toward unmet needs. One complaint doesn’t mean much. The same complaint appearing across different conversations deserves attention.
Save the wording customers use. It can help later when you’re comparing product features, interviewing buyers and explaining the product in language the customer already understands.
4. Watch Social Platforms for Changing Buying Behavior
TikTok, Instagram and Pinterest can reveal products or use cases before they show up clearly in traditional keyword data.
Look past view counts. Read the comments. See whether people ask where to buy the item, complain about the price or explain what they would change.
Workarounds deserve special attention. If people repeatedly modify a product to make it useful, the category may be missing a feature customers already know they want.
Then check whether the interest lasts. A product that appears in active conversations for months gives you a different signal than something driven by one viral video.
5. Study Crowdfunded Products
Kickstarter and Indiegogo show products that persuaded people to commit money before normal retail availability.
Study successful and unsuccessful campaigns in the same category. Look at what problem the product solves, how the creators explain the difference and what backers discuss in the comments.
Repeated crowdfunding activity can also show that buyers are open to new options in a category. Be careful with unusually large campaigns. A successful launch can reflect an existing audience, strong media coverage or a highly experienced team as much as the product itself.
Your goal is to understand what buyers responded to and what happened after the initial excitement.
6. Read Competitor Reviews Like Product Research
Competitors have already paid to answer one useful question: will somebody buy this kind of product?
Your job is to find out what customers still dislike after buying it.
Read positive, negative and middle-of-the-road reviews. Five-star reviews tell you what customers value. One-star reviews can expose failures. Three-star reviews are often useful because buyers explain what worked and what kept the product from being great.
Look for repeated complaints about construction, materials, sizing, setup, instructions, storage or packaging. Then ask whether fixing that issue would change the buying decision enough to support another product.
Order samples when the opportunity becomes serious. A complaint about a “weak latch” makes more sense when you can handle the product and see how the part was designed.
7. Use Keyword Data to Measure Search Interest
Keyword tools can show how often people search for a product, problem or use case. Google Keyword Planner, Semrush, Ahrefs and similar platforms can all help.
Don’t reduce the decision to one search-volume number. Look at the whole group of related searches.
A product might have modest volume for one exact term while attracting demand through dozens of related searches. The reverse can happen too. A high-volume phrase may be mostly informational and produce little evidence that people want to buy a physical product.
Compare search demand with marketplace activity. When people search for a problem and comparable products also sell, you have a stronger signal than either source gives you alone.
8. Talk to People Who Have the Problem
Customer conversations can explain what the numbers miss.
Ask people about what they already do rather than asking whether they’d buy your idea. Past behavior gives you better information than polite predictions.
Ask them to describe the last time the problem happened. Find out what they used, what bothered them and what they did next. Listen for workarounds, repeat purchases and products they’ve abandoned.
Don’t pitch while you’re interviewing. Once you explain your idea, people start reacting to your solution instead of describing their actual behavior.
After several conversations, compare the notes. A problem becomes more interesting when different customers describe it without being prompted.
How to Combine Conflicting Research Signals
Product research rarely gives you a clean yes or no. Different sources measure different behavior, so conflicting signals are normal.
Amazon sales tell you that people currently buy something. Google Trends shows changes in search interest. Reviews tell you where existing products disappoint. Customer conversations explain the reason behind the behavior.
Start by identifying which signal is most relevant to the product you’re considering.
Strong Sales With Weak Reviews
This can be interesting. Buyers are spending money, yet current options leave them unhappy.
Look for a complaint that appears across several competitors. Then decide whether you can fix it without pushing cost, complexity or selling price beyond what customers will accept.
Rising Interest With Little Marketplace Activity
You may be seeing an early category, a product people buy elsewhere or a topic that attracts interest without purchase intent.
Dig into the search terms and customer conversations before treating the trend as product demand. Find out what people expect to buy and where they currently buy it.
Large Market With One Strong Incumbent
A large market can still be a weak opportunity if customers already have a preferred choice and little reason to switch.
Study the customers the leading product serves poorly. An opportunity may exist around a specific audience, feature or price point. If you can’t explain the difference clearly, market size alone isn’t enough.
Enthusiastic Interviews With Weak Buying Evidence
People are generous with compliments when no money is involved.
Move the idea into a stronger test before spending heavily. Ask for an email signup, reservation, preorder or another action that requires more commitment than saying the idea sounds useful.
When signals disagree, document the conflict. Don’t average it away. The disagreement often tells you what needs to be tested next.
Compare Ideas With a Product Opportunity Scorecard
Once you’ve narrowed the field, score each product from 0 to 3. The scorecard gives you a consistent way to compare ideas without pretending every category follows the same benchmark.
0: No evidence
1: Weak evidence
2: Mixed evidence
3: Strong evidence
| Factor | Evidence to record |
|---|---|
| Demand | Search trend, comparable sales and repeat purchasing |
| Customer problem | Complaints repeated across reviews, forums and interviews |
| Difference | A feature, audience or experience competitors do not serve well |
| Unit economics | Estimated landed cost, fees, returns and customer acquisition cost |
| Feasibility | Supplier feedback, minimum order quantity, testing and lead time |
Record the source and date beside every score. A demand score based on last week’s Amazon activity means something different from a trend report you saved two years ago.
Compare products within the same category whenever possible. Acceptable pricing, margins, competition and order quantities can vary widely between a simple textile product and an electronic device.
The scorecard helps expose weak assumptions. Suppose an idea scores well on demand and customer complaints but poorly on economics. That tells you where to investigate before falling further in love with the idea.
Don’t turn the total into an automatic pass-or-fail number. Two products can earn the same total for very different reasons. A serious feasibility problem deserves more attention than a weak social-media signal.
Signs an Idea Needs More Work
Some problems should slow you down before you move into development.
Search interest and comparable sales are both falling.
You can’t find a repeated customer problem worth solving.
Your idea depends on being cheaper than established competitors.
Early cost estimates leave little room for fees, returns or customer acquisition.
Suppliers flag difficult tooling, testing or minimum order requirements that don’t fit your budget.
Customers like the idea in conversation but show little willingness to take a buying action.
One weak area doesn’t always kill an idea. It tells you what needs stronger evidence before you spend more.
Validate the Best Idea Before Manufacturing
Research helps you choose which idea deserves a real test. Validation asks buyers to do something that shows interest beyond a conversation or survey.
The right test depends on the product, price, sales channel and how familiar customers already are with the category. A landing page might work for one product. A preorder, reservation or prototype test may tell you more for another.
Avoid universal conversion targets. Traffic sources behave differently and a $20 impulse purchase shouldn’t be judged by the same standard as a $500 specialty product.
Set your pass-or-fail criteria before seeing the results. Decide what evidence would make you continue, revise the idea or stop. That keeps enthusiasm from changing the rules after the test.
Gembah’s guide to market validation covers these tests in more detail. Use that process once your discovery work has narrowed the field to the strongest opportunity.
Before paying for tooling or inventory, add one more check: production feasibility.
Ask suppliers whether the product can be made as planned. Get early feedback on materials, minimum order quantities, testing needs, tooling and likely lead times. At this stage, you’re looking for problems that could change the business case.
A promising market doesn’t help much if the product requires a production method your budget can’t support. Finding that out before detailed engineering gives you room to change direction.
What to Do Next
Take your strongest two or three ideas and put them through the same scorecard. Keep your sources, notes and dates beside each score so you can see why one opportunity looks stronger than another.
Then investigate the weakest assumption behind each finalist.
If demand is unclear, gather better sales or search evidence. If the customer problem is weak, conduct more interviews and review research. If the economics look tight, get better cost estimates. If production looks difficult, ask suppliers what would need to change.
Once one idea has enough evidence behind it, move into validation before committing to major development costs. If buyers respond and the economics still make sense, the next step is planning the product development process.
You should reach that stage with a clearer reason for building the product, evidence that customers care and a better idea of what could make the project expensive.
If you’re comparing product ideas and want a second set of eyes before committing to design or manufacturing, talk through the opportunity with Gembah. The goal is to find out which idea has enough evidence to deserve the next dollar you spend.
