LucernaSkin · At-a-glance brief

Lucerna.Skin

At-a-glance · Ads specialist intro
Prepared 2026-05-18 · Daniel Illingworth · Flowsly AI

DTC premium pezoneras for Spain/EU. Launched late 2025. Live Shopify storefront. Six months of paid testing has produced four sales total — the funnel is technically sound now, but the conversion problem is unresolved.

The brand in two lines

Medical-grade silicone gel nipple covers (8 cm, 0.5 mm matte, latex-free, ~30 uses) at €28.99 IVA-incl. Premium packaging (velvet bag + branded box + hygiene seal). Spanish-first, peninsular tuteo voice — calm, empathetic, no DR aggression. Live at lucernaskin.com (Shopify Dawn, apex Shopify-served since 2026-05-06).

Positioning: comparable price tier to Nippies/BOOMBA, premium of Nipskin. Closest analogue in your portfolio is Vanu (Ecuador) — same category, broadly similar AOV.

Unit economics — the margin envelope

Line	€
PVP (IVA incl.)	28.99
IVA 21%	−5.03
Net revenue	23.96
Product + packaging (cover, blister, caja, velvet bag, sticker, tarjeta)	−2.64
Shipping packaging	−0.30
Freight CN→ES (amortised over 350u)	−4.93
Card processing (2.1% + €0.30)	−0.91
Customer shipping (absorbed, free)	−6.00
Contribution margin / unit	€9.18 · 38.3%

At 1,000+ units freight drops ~60% → margin >45%. At 3,000u → 48%.

CAC ceiling €9.18 is the wall. Any CAC above this is underwater on the first sale.

CAC	Margin/u	Margin %
€5	€4.18	17.5%
€9.18	€0.00	—
€10	−€0.82	−3.4%
€15	−€5.82	−24.3%

Break-even (fixed costs €1,146/mo incl. ads + content): 125 orders/mo = €3,618. Post-content-contract: ~65/mo. Single-purchase product; LTV not modelled.

Inventory & exposure

350Units on hand

4Total sales to date *

€5,171Inventory at cost

€10,147Inventory at retail

€3,214Potential gross margin

* 4 Shopify orders Jan–Apr 2026 (€113.07). 2 attributable to Meta but landed in Shopify’s “unknown” bucket (fbclid lost on IG-profile bounce); zero Meta-attributed purchases in 90 d wrapper.

Meta ads — what’s been spent and learned

Account: act_1356533645384833 · EUR · Pixel 1239157417782558 · agency-run until 2026-04, in-house since.

Pre-baseline — agency era (2026-01-21 → 2026-04-20, 90 d)

€658Spend

231kImpressions

22.9kClicks · 9.9% CTR *

0Attributed purchases

7Campaigns (2 active)

* CTR mostly phantom. 94% of spend was OUTCOME_TRAFFIC with VIEW_INSTAGRAM_PROFILE CTAs — clicks landed on instagram.com/lucernaskin, not the store. Real click→LPV on the main IG-traffic campaign was 1.27% (8,197 → 104). Verdict: agency optimised for profile visits, not sales.

2026-04-22 baseline audit

Health 40/100 (D). P0s: wrong objectives, no CAPI, no custom audiences, ad-set names lying about targeting (“18-25” actually targets 18-65), identical 11-interest stack copy-pasted across every ad set. Catalog connected but no Advantage+ Sales campaign.

2026-04-28 → 05-01 — relaunch (Sales campaign, 3-day window)

Day	Spend	Impr	Clicks	LPV	VC	AtC	IC
04-29	€30.55	3,311	77	16	20	9	3
04-30	€22.69	3,022	69	30	35	2	1
05-01	€4.00	519	14	4	6	0	0
3-day	€57.24	6,852	160	51	62	11	4

30-day rollup: 5,344 LPV → 463 VC → 18 AtC → 19 IC → 0 Purchase. IC→Purchase 0/19 (industry norm 25–40%; expected ~5–8). Health rose to 56/100 (D+). CAPI relay + LDU consent gate verified live. Click→LPV recovered to 32% (vs 80%+ healthy — partly bot traffic, partly tracker-blocking browsers).
Root cause diagnosed: launch ads ran emoji/ALL-CAPS “VENTA FLASH 20% Dia de la Madre” urgency copy — categorical mismatch with calm editorial landing voice. Cookie modal hid the auto-applied discount above the fold on mobile. 12 ads rewritten in-brand for the next round.

2026-05-12 — current live launch

12 ads (4 concepts × 3 variants each, 1x1 + 9x16 formats) across two ad sets: LK_Sales_2026Q2_product_led (concepts 1–3: invisibility, no-adhesive centre, 30-use durability) and LK_Sales_2026Q2_summer_led (concept 4: summer occasions).
Landing: PDP at /products/cubrepezones, UTM-tagged per ad.
Voice: in-brand peninsular tuteo, no emojis, no ALL-CAPS — reverses the May 1 mismatch.
Funnel hygiene shipped same day: email popup + APROVECHA 10% inline offer suppressed for paid traffic (allowlist + click-ID + sessionStorage stickiness). Verified live on iPhone Safari.
Too recent for meaningful performance data in our snapshots — this is the dataset the specialist will be looking at fresh.

Data hygiene — what changed under the hood

CAPI relay live since ~2026-04-24 via in-house Next.js route + Shopify’s native Meta app for server Purchase. Pixel `1239157417782558` browser + server with shared eventId dedup, _fbp cookie, LDU on for unconsented visitors. Verified end-to-end live 2026-05-01.
Storefront cutover 2026-05-06: apex moved back to Shopify-served Dawn theme (Vercel headless preserved at guatitienda.vercel.app as preview/rollback).
Paid-traffic popup suppression 2026-05-12: removes a fold-blocking conversion killer for ad traffic specifically.
Still missing: zero custom audiences, zero lookalikes, zero customer-list upload. Single broad ad set (ES, 24–65). No retargeting layer — the largest unaddressed structural gap.

Honest framing for the specialist: the spec sheet says “DTC brand with 350u inventory, €9 contribution margin, 6 months of paid history, 4 lifetime sales.” The infrastructure is now sound (CAPI, sales objective, landing message-match, paid-traffic popup gate). The remaining unknown is whether the funnel can complete a purchase at all under brand-voice creative, or whether the price/proof gap (7 reviews on the PDP, €23 AOV cold) is the binding constraint.
Source docs: ops/projects/financials/unit_economics_summary.md · ops/reference/meta-ads-baseline/ads-meta-audit-output.md (Apr 22) · ops/reference/meta-ads-baseline/wrapper-summary-90d.md · ops/reference/meta-ads-baseline/2026-05-01-rebaseline/META-ADS-REPORT.md · .../2026-05-01-rebaseline/zero-sales-diagnosis.md · .../2026-05-12-landing-audit/meta-ads-paste-ledger.csv.