Social Voice-of-Customer — Bloomers Social

1,205 relevant Reddit records · 101 brand-mentioning · generated 2026-06-19

⚠ Separate bucket. This is ambient social/forum voice (Reddit), 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 Reddit conversation (0 of 105 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 101 recent brand-mentioning records (denominator = 105 total recent brand mentions; a comparison post counts once per brand). Trailing 18 months. Presence, not sentiment.

Knix
75.2%
n=79 recent · 114 all-time · CI 66%–83% · reddit
Soma
15.2%
n=16 recent · 64 all-time · CI 10%–23% · reddit
Jockey
5.7%
n=6 recent · 19 all-time · CI 3%–12% · reddit ≈ tied (overlapping confidence)
Thinx
2.9%
n=3 recent · 90 all-time · CI 1%–8% · reddit ≈ tied (overlapping confidence)
Bali
0.9%
n=1 recent · 2 all-time · CI 0%–5% · reddit ≈ tied (overlapping confidence)
Bloomers ← CLIENT
0.0%
n=0 recent · 0 all-time · CI 0%–4% · reddit ≈ 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 43% of captured Reddit records fall in the trailing 18 months (1,205 of 2,792). 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

92% of relevant posts (1,104 of 1,205) 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).

35% 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 this23% of posts · 0% negative · n=283
Menopause Perimenopause Incontinence Fit category21% of posts · 62% negative · n=255
Comfort Irritation product15% of posts · 69% negative · n=186
Style Aesthetic Coverage category11% of posts · 31% negative · n=130
Sizing Fit Advice category10% of posts · 33% negative · n=116
Comparison Vs Competitor product7% of posts · 23% negative · n=88
Where To Buy categoryreviews miss this7% of posts · 17% negative · n=83
Price Value Seeking categoryreviews miss this7% of posts · 51% negative · n=81
Materials Fabric Preference category6% of posts · 12% negative · n=74
Durability Quality Decline product4% of posts · 59% negative · n=44
Health Safety Concern productreviews miss this3% of posts · 88% negative · n=42
Leak Absorbency product3% of posts · 68% negative · n=34
Fitting Service Retail Experience category3% of posts · 55% negative · n=33
Dupe Or Alternative Seeking categoryreviews miss this1% of posts · 15% negative · n=13
Sustainability Ethics categoryreviews miss this0% of posts · 25% negative · n=4

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 5.9% 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.
Aspect2023-Q12023-Q22025-Q12025-Q22025-Q32025-Q42026-Q12026-Q2
Menopause Perimenopause Incontinence Fit5%14%19%20%23%15%23%24%
Recommendation Request Social Proof24%1%16%19%26%38%16%22%
Comfort Irritation56%2%26%13%14%22%12%11%
Price Value Seeking10%1%3%5%4%11%4%10%
Comparison Vs Competitor5%—6%5%10%10%4%8%
Sizing Fit Advice10%1%9%10%11%13%7%7%
Where To Buy7%1%8%6%4%10%6%7%
Style Aesthetic Coverage46%1%28%8%7%19%6%5%
Health Safety Concern5%—6%1%4%2%3%5%
Fitting Service Retail Experience——3%3%3%1%2%4%

Head-to-heads — where buyers compare brands (14)

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

JockeySoma
KnixThinx
KnixThinx
KnixThinx
KnixSoma
BaliJockeySoma
KnixThinx
JockeySoma
BaliSoma
KnixSoma

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=79
Soma n=16
Jockey n=6
Thinx n=3
Bali n=1
Bloomers 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. Reddit 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.