Researching Market Trends from SERP
Analyzes cluttered search engine results (mixed Etsy listings, Pinterest pins, TikTok/Instagram posts, Reddit threads) to surface concrete product ideas, naming patterns, and pricing signals for a craft maker deciding what to create next.
Given a raw SERP dump about "polymer clay Christmas stud earrings," scan for:
- Repeated subject motifs (what characters/objects keep appearing across listings)
- Price anchors (Etsy prices found in the dump)
- Format patterns (studs vs. dangles, sets vs. singles, cutters vs. hand-sculpted)
- Gaps — subjects the asker already made that are underrepresented relative to what's trending
Output a ranked shortlist of 8-12 new product ideas, grouped by theme, with one-line rationale citing what was seen in the results.
Progress:
- [ ] Step 1: Strip source noise (site names, "Etsy·$X", "Pinterest", "TikTok", video tags)
- [ ] Step 2: Extract every distinct subject/motif mentioned
- [ ] Step 3: Tally frequency — which motifs appear 3+ times vs. once
- [ ] Step 4: Note format details (stud vs dangle, set size, cutter-based vs sculpted)
- [ ] Step 5: Pull price data points where present
- [ ] Step 6: Cross-reference against what the user already made (don't re-suggest duplicates)
- [ ] Step 7: Produce ranked idea list with rationale + 2-3 "stretch" novel ideas not seen verbatim but adjacent
Step 1 — Strip noise: Remove platform labels (Etsy, Pinterest, Facebook, TikTok, YouTube, Instagram, Amazon, Shein), prices in isolation, and "Video"/"days ago" timestamps. Keep only the descriptive title/snippet text.
Step 2 — Extract motifs: Read every listing title for the actual subject (e.g., "gingerbread man," "nutcracker," "candy cane," "holly," "pine cone," "igloo," "fireplace," "snowflake," "cookie tray," "Grinch").
Step 3 — Tally: Count mentions. High-frequency = validated demand (safe bets). Low-frequency/unique = differentiation opportunities.
Step 4 — Format patterns: Note whether items are sold as studs, dangles, "build your own pack," sets of 3/5/6, or paired with trays/magnets/charms rather than earrings only.
Step 5 — Pricing: Collect any $ figures to give a market price range.
Step 6 — Dedupe against existing work: Exclude motifs the user states they already made.
Step 7 — Final list: Present as categorized, prioritized suggestions.
Example 1: Input: SERP dump for "polymer clay Christmas stud earrings" where user already made snowman, Rudolph, Christmas tree. Output:
Already covered: Snowman, Rudolph, Christmas tree
High-confidence additions (appeared 3+ times across results):
- Gingerbread man / gingerbread house — strong repeat presence, pairs well with "cute miniature food" niche
- Candy cane — frequently bundled in sets
- Holly / pine cone — common filler pieces in "build your own pack" listings
- Santa (hat or full figure) — appears across multiple shops as a core set piece
- Penguin / bear (winter animals, non-Christmas-specific) — seen in KEOKER's bestselling sets, extends sell season
Differentiator ideas (lower frequency, stand out):
- Nutcracker — seen once, underused, visually distinct
- Cookie platter / cookie tray earrings — novel trend (multiple Etsy shops), could be a signature set
- Fireplace / igloo — niche "cozy winter" subset
- Christmas book stud — unusual, appeared once, good for "cozy reading" crossover audience
Format notes:
- Sets of 5-6 studs sell well as "build your own pack" (Etsy, $14-$28 range)
- Mixing a non-literal theme (winter animals, cozy cabin) extends shelf life past Dec 25
Price benchmark: $8.60 (magnets) to $27.98 (stud sets) — most stud sets cluster $14-$26.
Example 2: Input: SERP dump mentions "pretzel earrings," "Stardew Valley charms," "dachshund holiday set" as one-off maker posts. Output: Flag these as "novelty/non-obvious" ideas — low competition, appeal to niche hobby/food crossover audiences; worth testing as limited one-offs rather than core line.
- Prioritize motifs appearing across multiple different sellers/platforms, not just one listing — that signals real demand, not a single maker's quirk.
- Separate "safe bet" (high frequency, proven) from "differentiator" (rare, distinctive) so the maker can balance a collection.
- Always exclude what the user already has to avoid redundant suggestions.
- Surface format/bundling patterns (sets, build-your-own packs) — these often matter as much as subject matter.
- Include pricing data when present; it's actionable for the seller.
- Note seasonal-extension opportunities (e.g., winter animals that sell past Christmas).
- Don't just list every noun seen once — that's noise, not signal. Require repetition or strong platform diversity before calling something "trending."
- Don't ignore non-Etsy sources (TikTok/Instagram captions) — they often reveal emerging ideas before they hit marketplaces.
- Don't recommend items the user explicitly already made.
- Don't present raw scraped junk (site names, timestamps) as if it were content.
- Avoid generic suggestions ("make more animals") — always tie back to specific motifs actually observed in the data.