The five-step method
Step 1: Build an idea list from problems
Write down 50 problems people pay to solve. Good places to look:
- Amazon "Movers & Shakers" and bestseller lists in categories you understand
- Reddit and forums for hobbies or groups you know: look for "what do you use for…" and "I wish there was…" posts
- Google autocomplete: type "how to stop", "best way to", "[hobby] for beginners" and note the suggestions
- Your own life: things you bought in the last year to fix an annoyance
Write each idea as a search phrase a shopper would type: "posture corrector for women", "dog car seat cover", "under sink organizer".
Step 2: Check the trend shape
Open Google Trends, set the location to your selling country and the time range to Past 5 years. Enter up to five ideas at once. You're looking at the shape of the line, not the number.
Step 3: Check search volume
Open Google Ads → Tools → Keyword Planner → Discover new keywords. Enter your idea and the selling country. (You need a Google Ads account, but you don't have to run ads.)
- Monthly searches. As a rough guide for a single-product test, look for a few thousand monthly searches across the main phrase and close variations. New accounts see ranges, like 1K–10K. That's enough to compare ideas.
- Top of page bid. This shows roughly what advertisers pay per click. If the high range is more than about 5% of your planned selling price, ads will be expensive.
- Related phrases. Lots of specific long phrases ("waterproof dog car seat cover for SUV") is a good sign. It means shoppers know what they want.
Step 4: Look at the competition on Google Shopping
Search each remaining idea on Google and click the Shopping tab. Note:
| What you see | What it means |
|---|---|
| Amazon, Walmart and big brands fill every slot | Hard to win on price. Only go ahead if you have a clear difference. |
| Mix of small stores and some big ones | Good. There's room. |
| Same item from many stores, same supplier photos | Price war. Skip unless you can improve the product or bundle. |
| Prices mostly $30–$100 | Good range for paid ads to work |
| Prices mostly under $20 | Ad costs will eat the margin |
Step 5: Read the bad reviews
Find the top 2–3 competing products on Amazon. Filter reviews to 1, 2 and 3 stars. Write down every complaint that appears more than twice. Common ones: "broke after a month", "smaller than expected", "doesn't fit [specific model]", "instructions useless".
Each repeated complaint is a chance. You can choose a better supplier, add the missing size, include a clear guide, or say on your product page exactly how yours solves it.
Score each idea
Give each remaining idea 1–5 points in six areas. Test the two or three highest totals.
If the price is three times what the product costs to get to the customer, you have roughly two-thirds left to cover ads, fees and returns. Below that, paid ads rarely leave a profit. See the break-even maths in our small budget Shopping guide.
Google E-commerce Product Research Method 2026
The full scoring sheet, keyword research walkthroughs by niche, and how to run a low-cost Shopping test to confirm demand before you buy stock.
[ Open the course ]