What if you do not know what to search?
You may know how to work with wood, sew small accessories or pack fragile ceramics without knowing the language shoppers use. Discovery gives you a place to begin. Open the niche feed and inspect candidate ideas before choosing an exact product query.
The initial categories include wooden home decor, desk organization, pet accessories, wedding keepsakes, kitchen storage and jewelry storage. These are starting directions, not a list of proven bestsellers. Ideas from saved listing titles add more specific products as evidence becomes available.
How a starting idea becomes research
- Find a product phrase. The engine looks for recurring phrases across distinct saved listings and retains example titles and their source query. Repeated captures of one listing do not become independent support.
- Read where it came from. A broad starter category and a phrase found in collected titles have different origins. Inspect that origin before treating the idea as a product direction.
- Research the exact phrase. When discovery is enabled and Chrome is available, the engine queues browser searches for candidates. Evidence from a parent query does not substitute for the candidate’s own search.
- Review the next step. Open researched listings, keep investigating a relevant idea or dismiss one that does not fit. Pause discovery when you want to stop new research; saved ideas and history remain available.
The discovery feed is part of the early-access app. Keep the app service and connected Chrome running for new collection. “Refresh saved ideas” reuses existing observations; it does not request another Etsy page.
Keep the product intact: a watch box example
Qualitative example, not measured market data. Suppose saved titles contain “wooden watch box” and descriptions of watch storage. This could suggest a box-making direction worth opening. It does not establish how many buyers search for it.
| Candidate phrase | Product meaning to check | Useful next check |
|---|---|---|
| wooden watch box | A box for storing watches. | Open photos and details: compartments, cushions, lid and overall dimensions. |
| wood watch | A wooden watch or another watch-related product. | Check the actual object. Exclude wearable watches when researching storage boxes. |
| wooden box | A broad container category. | Separate watch storage from packaging, keepsakes and craft supplies. |
The engine filters obvious incomplete fragments and promotional wording, but you still review meaning. Prefer a candidate you can describe as an object plus a use. “Gift for him” alone does not tell a maker what to produce.
A sensible branch to explore might be a single-watch case versus a multi-watch box. Research each phrase separately and compare like-for-like products. Do not inherit the broad category’s observations as evidence for the narrower idea.
Read the idea’s state before its evidence
| What you see | What it means | What to do |
|---|---|---|
| Starting category or unresearched idea | A direction is available, but its own search evidence is not ready. | Inspect the idea’s origin and whether you could make the product. |
| Researched idea | The app has a query-specific sample with a recorded time. | Open the listings and check product relevance, sample coverage and missing fields. |
| Waiting for Chrome or collection paused | New pages cannot currently be read. | Check the connection or pause reason; do not interpret waiting as an empty market. |
| Older or partial reading | The stored evidence has an age or a coverage limit. | Revisit the original time and collected pages before making a production decision. |
A dash means unavailable, not zero demand or zero competition. Title repetition does not measure search volume. An absent cart message does not measure zero purchases. Classification percentages exclude unknown rows and may remain unavailable when too few values are readable.
Turn discovery into a shortlist you can explain
Choose a small set of product directions that fit your materials, equipment, packaging and delivery constraints. For each, write one sentence about the buyer’s use, save relevant listing examples, note mismatches and identify the most important unanswered question.
Use product research for the exact-query comparison, then fill in the research worksheet. A candidate can be worth a prototype, need another reading, or be unsuitable for your business. Keeping the reason is more useful than treating a feed position as a promise of success.
Our discovery method
Read our research methodology: data sources, coverage and missing values.
Written by the NicheWright team. Candidate generation uses broad starters and recurring phrases from saved titles; browser collection supplies separate query-specific evidence. Example phrases on this page explain that distinction and do not report a live scan.
Etsy’s search guidance describes shopper-specific results. A candidate’s captured listings therefore provide context for that session, not a universal ranking or a market-wide demand reading. Source: Etsy’s How Etsy Search Works, reviewed September 6, 2026.