**To:** Freya Lindgren
**From:** Zara Ahmed
**Re:** Cragsley, the December dip, images, and bundles — what I found

**1. Cragsley's quality problem is specific, recurring, and currently quiet rather than urgent.** (*Poor Ratings by Category*, *Rating Trend*, *Sales Trajectory*) Poor ratings concentrate in soft goods — Loungewear (66.7%/40.0% of reviewed products rated <3), Outerwear (50.0%/28.6%), Shirts (50.0%) — while Boots, Knitwear, and Scarves are clean (0-17%). It's not a new crisis: December 1-2-star share has spiked every year (25.0% in 2023, 22.1% in 2024, 25.7% in 2025) alongside the seasonal order surge. The poor-rated products' revenue grew in absolute terms (€896→€6,177/month) but their *share* of Cragsley's total fell from ~23% to ~6% as the rest of the catalog scaled much faster. Read: reputation is quietly eroding relative to the business, not crashing it. **High confidence** — large sample (105 reviewed products, 36 months of data).

**2. The December dip had two different causes, not one.** (*Fennel & Flint / Wiesgrund Wohnen: Delivery vs Reviews*, *Historical December Pattern*, *Product Breakdowns*, *Fragile Category Returns*) Fennel & Flint shows a genuine 3-year decline (Dec avg rating 4.43→3.71→3.14) as orders grew 28→45→69, worsened by a real November delivery breakdown (on-time delivery fell to 53.7% from ~85%) that likely fed December's reviews with a one-month lag — this is a scaling/ops problem, not a single bad product (14 products affected, no clean cluster). Wiesgrund Wohnen is different: delivery stayed strong (93.8% on-time, 0% via the troubled Albion Parcel carrier) and the 3-year trend is noisy, not declining (4.00→3.57→3.50). Their December reviews were thin (1-2 each) but landed overwhelmingly on fragile homeware — vases, plates, a casserole dish. Platform-wide, Home & Living's fragile categories do show higher damage-related return shares (Decor 26.5%, Tableware 30.0%, Glassware 26.7%, Furniture 40.0%) versus sturdier goods (Bedding 3.6%, Towels 3.7%) — consistent with in-transit breakage rather than a quality or delivery-speed issue. **Medium confidence** — Wiesgrund's own sample is small (10-12 reviews); the fragile-category pattern is suggestive, not proven for this specific merchant.

**3. The zero-to-one-image jump looks like a real effect, not a stand-in.** (*Image Buckets: Age, Status, Vertical*) The 85 zero-image products are all currently active (not draft/archived) and are actually the *oldest* listings on the platform (1,136 days old vs. ~1,040 for others), with fewer variants (3.4 vs. ~5.5) — neglected old listings, not unfinished new ones. No B2B merchant has a single zero-image product. This rules out my main alternative explanations. **High confidence**, though the group is small (85 products) — this is a cleanup task, not a trend.

**4. Bundle adoption is a genuine, stark vertical split — not a data artifact.** (*Bundle Adoption by Vertical*, *Bundle Creation Timing*) 100% of merchants in Beauty, Books & Stationery, Consumer Electronics, Food & Coffee, Home & Living, and Jewelry have made at least one bundle; 0% in Apparel & Footwear, both B2B verticals, Pet Supplies, and Sports & Outdoor have made any. I checked whether this was a seeding artifact — creation dates spread steadily over 3 years across many different tenants (e.g., 4 new tenants bundling in May 2025 alone), so it's organic. Depth stays shallow everywhere (2-5% of catalog bundled even among adopters). **High confidence** on the pattern; reads as "bundles fit giftable/discrete-item merchandising, not apparel sizing or B2B bulk."

**Caveats:** Small sample sizes drive several merchant-level percentages (Wiesgrund's 10-12 reviews, some category return counts as low as 10). Cragsley's category list had duplicate names, suggesting a minor categorization data-quality issue I haven't chased down.

**Next steps:** Watch Fennel & Flint's January ratings for continued decline; flag packaging/protective materials for fragile Home & Living categories with ops; audit the 85 zero-image listings for a quick content fix; and use the bundle-vertical pattern to target Beauty/Food & Coffee merchants for deeper (not just any) bundle adoption.