- Men’s & Women’s Fashion
- PROSPECT
- NEW · every 48h
- 1/day
- https://rileyriver.co/
Phase 2 — Research
Read the market before touching the account: what Pinterest rewards, who the competitors are, what the brand actually stands for.
- 14
- done
- 14
- in total
Step 2.1 · Grid & competitors
- What
- Pick five to ten seed keywords, search each one incognito, record what format dominates and how page one feels, extract the three dominant hex codes, then identify five to ten real competitors and export their pins.
- Where
- Pinterest search in an incognito window (or PinClicks to avoid personalisation). PinInspector for the competitor exports. Recording happens here.
- Output
- One grid record per keyword with format + CTA + overlay bucket + colours, plus competitor rows with their pin exports imported.
Choose seed keywords
Five to ten broad terms that describe the brand. Starting point for the grid analysis.
Seed keywords: mens fashion, womens fashion, mens clothing, womens clothing, mens shoes, womens shoes, mens outfits, womens outfits, mens style, womens style
Review the grid per keyword
Search incognito, or use PinClicks to avoid personalisation bias. Look at the first 15 to 20 organic pins. This reads what Pinterest already rewards for the keyword, so it has nothing to do with whether the store itself has any history — it applies to a brand-new account exactly as it does to a takeover.
Nothing here — the next task (P2.1.3) is the form this feeds. Look at the first fifteen to twenty organic pins per seed keyword, incognito so personalisation does not colour it, then record what you saw there.
Record the grid
Per keyword: which format dominates, are there CTAs, how page 1 feels. Structured input so that everyone delivers the same.
Grid recorded for 10 keyword(s).
Extract dominant colours
Right-click next to the top 3 pins, View Page Source, search for dominant_color. These hex codes drive the design brief.
Hex colors recorded for 10 keyword(s).
Identify competitors
Real brands with an active account in the same niche. Not Shein, not individual collectors.
Competitors: Banana Republic, Abercrombie & Fitch, Everlane, MANGO, Old Navy
Export competitor pins
Seven hundred to a thousand pins per competitor where the competitor has them, and everything they do have where they do not — the count per competitor is itself a finding about the niche. Five to ten competitors. Manual, do not automate.
The parsed exports, imported per competitor on the upload screen above: 700-1000 pins each where that many exist, otherwise everything the competitor has. Five to ten competitors — the coverage line tells you which are still missing, and the imported count per competitor is recorded with them. Attach the link to wherever the raw CSVs live (Drive) so the export can be traced back later; it lands in the Assets library as a PinInspector export automatically.
Imported 1,298 competitor pins across 5/5 competitors.
Collect top pin designs
Per keyword: pin URL, title, description, annotations and colours. Material for the AI analysis.
Per keyword: the pin URL, its title and description, the annotations PinClicks returned and the three dominant colours. This is what the AI market analysis reads — an annotation stays research until it passes a volume check.
Top pin designs recorded for 4 pin(s).
Step 2.2 · AI market analysis
- What
- The system assembles a prompt from the intake, brand data, grid analyses and competitor exports, and returns a Steal List, a Board Gap list and content angles. You approve or reject each item individually.
- Where
- In this dashboard. Nothing to paste — the prompt builds itself from what is already stored.
- Output
- Approved Steal List and Board Gap items mirrored into board opportunities for phase 3.
Run AI market analysis
The prompt is assembled server-side from the intake questionnaire, the taste graph, the grid analyses and the competitor list — there is nothing to paste and nothing to attach. Output is a Steal List, board gaps and content angles, each approvable one by one in P2.2.2.
AI market analysis generated 21 items.
Review insights
Pay particular attention to the Steal List: board names competitors use that we should adopt. And the Board Gap: what nobody is covering.
Reviewed market analysis: 17 approved, 4 rejected
Step 2.3 · Taste graph
- What
- Map seven fields (products, spaces, aesthetics, moments, functional outcome, aspirational outcome, related interests), read Audience Insights, then distil exactly three content angles, three visual worlds and three key moments.
- Where
- This dashboard, with Pinterest Audience Insights open alongside.
- Output
- A taste graph row and the three-by-three distillation that feeds board names and image prompts later.
Taste graph mapping
Seven fields: products, spaces, aesthetics, moments, functional outcome, aspirational outcome, related interests.
Taste graph mapped (7 fields).
Read Audience Insights
Affinities of the engaged audience. Surprising correlations often produce the best content angles.
The affinities of the engaged audience, with the surprising ones marked. A predictable affinity confirms what you already knew; a surprising one is where a content angle comes from.
10 affinities recorded, 8 marked surprising.
Three angles, worlds and moments
Distillation of the taste graph. Feeds board names, clusters and AI image prompts.
3 angles / 3 worlds / 3 moments distilled.
Step 2.4 · Frequency
- What
- Record how many pins per day the top competitors publish over four months, then let the system compute how many URLs per month the plan needs using the sixteen-pin math.
- Where
- Competitor velocity is counted manually from their profiles. The URL requirement is computed here.
- Output
- Competitor velocity per profile and a stored urls_per_month value on client settings.
Determine competitor velocity
Pins per day of the top 5, measured over 4 months. Sets the market standard you have to match.
Velocity recorded for 5 competitor(s).
Set frequency
The 16-pin math: 1 URL yields 16 pins. The system calculates how many URLs you need. Two per day for a year beats ten per day for a week.
1 pins/day × 30 days ÷ 16 pins/URL = 1.88 → 1 URLs/month (rounded down to what the daily cap can publish)