Social Voice-of-Customer — Bloomers Youtube

597 relevant YouTube records · 9 brand-mentioning · generated 2026-06-19

⚠ Separate bucket. This is ambient social/forum voice (YouTube), a fundamentally different speech genre from solicited reviews — it is never ranked against the review report. Share-of-Voice measures presence, not sentiment. Brand attribution is in-text with abstain; single-mention attributions are within attribution noise.

Share of Voice — who owns the recent organic conversation

0.0%
Bloomers is absent from the recent organic YouTube conversation (0 of 11 brand mentions in the trailing 18 months). Competitors own the word-of-mouth in the communities where this category is decided.

Share of the brand mentions across 9 recent brand-mentioning records (denominator = 11 total recent brand mentions; a comparison post counts once per brand). Trailing 18 months. Presence, not sentiment.

Knix
54.5%
n=6 recent · 29 all-time · CI 28%–79% · youtube
Thinx
36.4%
n=4 recent · 31 all-time · CI 15%–65% · youtube ≈ tied (overlapping confidence)
Soma
9.1%
n=1 recent · 17 all-time · CI 2%–38% · youtube ≈ tied (overlapping confidence)
Bloomers ← CLIENT
0.0%
n=0 recent · 0 all-time · CI 0%–26% · youtube ≈ tied (overlapping confidence)
Bali
0.0%
n=0 recent · 0 all-time · CI 0%–26% · youtube ≈ tied (overlapping confidence)
Jockey
0.0%
n=0 recent · 3 all-time · CI 0%–26% · youtube ≈ tied (overlapping confidence)
⚠ Capture-recall caveat. Share-of-Voice measures who we captured more of — not necessarily who is discussed more. If our capture recall differs by brand, this over- or under-states a brand's true conversation share; the confidence band reflects sampling noise only, not capture bias.
⚠ Thin window. Only 38% of captured YouTube records fall in the trailing 18 months (597 of 1,569). Consider widening the window globally (a named alternative-window build) or treating the thin brands as baseline only — never widen per brand.

Category discourse — the bigger story

98% of relevant posts (588 of 597) name no brand at all — pure category conversation (menopause/perimenopause fit, leak protection, dupe hunts). This is the unmet-needs space buyers discuss before choosing a brand.

Category-pain map — what the community actually discusses

Aspect prevalence across relevant posts in the reporting window (the 15-aspect social taxonomy, discovered from this corpus). category = problem-space / shopping behaviour; product = product experience. Click to read posts (evidence highlighted).

45% of aspect mentions are social-only themes a review funnel structurally can't capture (where-to-buy, dupe-hunting, social proof, price-seeking, sustainability). This is the differentiated intelligence in the ambient bucket.
Recommendation Request Social Proof categoryreviews miss this14% of posts · 2% negative · n=81
Style Aesthetic Coverage category8% of posts · 28% negative · n=50
Comfort Irritation product8% of posts · 43% negative · n=47
Leak Absorbency product7% of posts · 36% negative · n=42
Price Value Seeking categoryreviews miss this7% of posts · 41% negative · n=41
Where To Buy categoryreviews miss this7% of posts · 15% negative · n=40
Sizing Fit Advice category5% of posts · 40% negative · n=30
Materials Fabric Preference category5% of posts · 23% negative · n=30
Health Safety Concern productreviews miss this4% of posts · 72% negative · n=25
Comparison Vs Competitor product4% of posts · 17% negative · n=24
Durability Quality Decline product3% of posts · 50% negative · n=16
Sustainability Ethics categoryreviews miss this2% of posts · 23% negative · n=13
Menopause Perimenopause Incontinence Fit category2% of posts · 0% negative · n=9
Dupe Or Alternative Seeking categoryreviews miss this0% of posts · 0% negative · n=2

Topic-mix over time — how the conversation evolved

Governed within-platform: a seed-stability check (scripts/seed_stability_social.py) confirmed the topic-mix is robust to which brands were used as capture seeds (max drift 2.5% under seed hold-out, below threshold). Share of captured discourse (conditional on collection method), per quarter. Within-period rates (never raw counts); only quarters above the evidence floor are shown. Never co-plotted against the review topic-mix — a different genre.
Aspect2021-Q22022-Q22023-Q22024-Q22025-Q22025-Q32026-Q12026-Q2
Recommendation Request Social Proof18%17%13%16%11%12%19%15%
Materials Fabric Preference2%3%1%7%5%2%1%14%
Comfort Irritation5%8%7%7%7%3%12%12%
Leak Absorbency13%15%4%8%7%3%6%10%
Style Aesthetic Coverage3%6%9%5%3%13%19%10%
Health Safety Concern4%2%2%5%3%3%4%9%
Where To Buy3%5%4%6%4%11%9%8%
Comparison Vs Competitor5%5%1%4%3%—9%6%
Price Value Seeking6%6%7%4%3%11%10%6%
Sizing Fit Advice3%7%3%4%3%2%15%4%

Head-to-heads — where buyers compare brands (4)

Posts that weigh multiple brands against each other, most-upvoted first. Click to read.

KnixThinx
KnixThinx
KnixThinx
KnixThinx

By brand — read the mentions

Most-upvoted (salience-ranked) attributed mentions per brand; the evidence span that drove attribution is highlighted. Click to expand.

Knix n=6
Thinx n=4
Soma n=1
Bloomers n=0 — no recent mentions
Bali n=0 — no recent mentions
Jockey n=0 — no recent mentions
How to read this. Every number is computed in SQL over the governed metric_social_sov / social_brand tables (the model never generates figures); each brand expands to the verbatim posts (evidence span highlighted) behind it, linked to the source thread.
Window & lanes. Headline comparison = trailing 18 months; full history is retained (excluded from the comparison only, never deleted).
Limitations. YouTube capture was brand-seeded (the captured SET is a seed artifact); comment context is not yet used (standalone classification); SOV is presence-only. Single-mention attributions are within attribution noise.