👥 Team Meeting6
Team MTGteam2026-07-10
Reset the influencer program and pivot the whole shop from mass output → research-first quality. Huma re-contacts ghosted influencers under the alias "Hannah" — apologize (Tina left), promise timely comms, offer free jelly sticks to rebuild trust. New rule: prioritize engagement (comment activity) over follower count — ~$80/package JP shipping makes mass-sending unsustainable. Isabella = top ambassador target (~500 comments/post, already uses Miranest) → propose a paid $300–$500/mo, 3–6 month deal to license her content for ads + landing pages. All new creative must be built on proven research (competitor breakdowns), not raw ideas. Germany is bleeding (1 sale in 3 days) → test three new 10% lookalikes (Pixel / Purchase-180 / Shopify list), research German shopping habits, and lean into educational content (out-converting before/after in the US).
Decisions
- Influencer recovery play: Huma re-contacts ghosted influencers as "Hannah" — apologize + explain Tina's departure, promise timely communication, and offer compensation (free jelly sticks) to repair the relationship.
- Quality over quantity: stop mass-producing ads/influencer sends. The ~$80/package shipping cost from Japan makes volume unsustainable — prioritize influencers by engagement/comment activity, not follower count.
- Isabella → ambassador: highest-value target (~500 comments/post, uses Miranest, lifestyle fit). Propose a paid $300–$500/mo, 3–6 month partnership to use her content in paid ads + landing pages.
- Research-first creative: every new ad must be grounded in proven concepts (competitor analysis / ad breakdowns) with a written rationale — master winning-ad psychology (hooks, structure), then apply the team's editing strength.
- Germany fix: market is underperforming (1 sale in 3 days). Test three new 10% lookalikes (Pixel, Purchase-180, Shopify list) to lift spend; research German shopping/review habits; pivot toward educational content (US: 12 conv from 8 educational videos vs 15 from 37 before/after ads).
Action items — tick when done
Team MTGteam2026-07-09
Fix Miranest Germany's poor CPO — root cause is self-competition (our agency + PSJN running similar ads on the shared Miranest Global page/pixel, inflating auction prices). Stop current campaigns and test two new DE campaigns, 25k JPY each, 7-day read: (1) a 1% purchase LLA from a combined US+DE customer list, (2) broad targeting on a brand-new Facebook page to reset delivery. Upgrade influencer vetting (judge comment purchase intent, not vanity engagement). Shift creative to educational animation + storytelling ("Lazial" / "Swiftlet Revolution" style). New hire Mao (17, Japan) joins to learn digital marketing.
Decisions
- DE fix = two-campaign test. Root cause: two agencies (ours + PSJN) sharing the Miranest Global page/pixel confuse Meta's algorithm → higher auction prices, weaker delivery. Stop current campaigns; launch two at 25k JPY each (50k total), 7-day read — (1) 1% purchase LLA from a combined US+DE buyer list (leans on proven US pixel data to find high-intent DE buyers), (2) new Facebook page on broad targeting to reset delivery.
- Influencer vetting upgraded. Paid collabs: judge comment quality (product questions = purchase intent) over vanity metrics + review past brand-post sales. Free collabs: confirm responsiveness before shipping product + enforce brand fit (Bisu is strict). Lock all terms up front. Cross-promote winning Miranest influencers → Bisu and vice-versa.
- Creative pivot to storytelling. Move off transactional "it works, buy now" → educational animation mixed with real footage; benchmarks = the "Lazial" animated sialic-acid explainer + the "Swiftlet Revolution" brand film; align to the new DE hair-loss angle.
- New team member: Mao (17, from Japan) joins to learn digital marketing; include in strategy meetings so it informs the creative work.
Action items — tick when done
Team mtg: july 6 weekly to do list with teamteam2026-07-06
Shift ad strategy to quality over quantity — 80% deep research (viral hooks + proven templates), 15% ideation, 5% AI-assisted editing — before mass production. Influencer outreach is the top priority (Thu client presentation). For Germany: pivot messaging from "Swallow's Nest" → "superfood", target the German Spitz breed (Pomeranian before/afters as proxy), and replicate proven US templates F2/F45.
Decisions
- Quality over quantity: stop mass-producing ads (~60/wk, weak results); split effort 80% research / 15% ideation / 5% editing (AI tools for basic edits), build a repeatable system, then scale.
- German messaging pivot: replace "Swallow's Nest" (flagged by the German agency) with "superfood," "ancient superfood," or "cellular renewal jelly stick."
- German creative focus: target the German Spitz breed; use Pomeranian before/after photos as a proxy (closely related breeds).
- Replicate proven templates F2 & F45 (top US performers) for the German market.
- Influencer outreach = top priority for Kim & Huma, driven by a client presentation this Thursday.
Action items — tick when done
Team mtg - how to use the hk agency softwareteam2026-07-06
Walkthrough of the new HK Agency platform — the team's central hub for performance tracking, to-do lists, meeting summaries and idea sharing. Judge daily performance on Meta CPO (🟢 <¥25k · 🟡 ¥25–35k · 🔴 >¥35k); for video ads chase a >25% hook rate (viewers held through the first 3s). Weekly rhythm: build a personal to-do list + check meeting summaries for team action items.
Decisions
- Judge daily performance on Meta CPO, color-coded 🟢 <¥25k · 🟡 ¥25–35k · 🔴 >¥35k; use the 7-day view to smooth daily volatility.
- Video-ad north star = >25% hook rate (held through first 3s); a low hook (e.g. 18%) = weak opening to fix. Hold rate = % who watch the whole video.
- Weekly workflow: each member starts the week with a personal to-do list + checks meeting summaries for team-wide action items.
- Diversify creative by style — data shows a heavy UGC bias; deliberately test underrepresented styles (story, problem) and iterate winners (e.g. Alopecia → Alopecia 2, 3).
Action items — tick when done
Han AI video & New business ideateam2026-07-01
Revenue dropped ~75% (single-client dependence) → agency now at a loss. Two moves: (1) fix underperforming AI ads by mimicking winning UGC frame-for-frame, and double ad spend next month to hit break-even; (2) de-risk by launching a proprietary AI-enhanced digital product/service — shift from service agency → product company.
Decisions
- AI ads: stop guessing — clone successful UGC videos frame-for-frame as the realism benchmark. (An ad Meta won't spend on = a losing ad; kill fast.)
- Cash: double ad spend next month to reach break-even.
- Strategy: build a proprietary product/service (digital / AI-enhanced, not physical) for a stable second revenue stream. North-star example: competitor's $150 AI beauty report (was ~$10k in-clinic) reportedly doing $2M/mo.
Action items — tick when done
Influencer CP Guide (Kim & Fuma)team2026-06-30
Past influencer campaigns made posts but didn't convert. Shift: treat influencers as content creators for ads, not social posters. Two tiers — free (volume, commission) + paid (quality, flat fee + product) — with strict vetting. $3,000/mo test budget across Bisu/Miranest.
Decisions
- Reframe influencers as ad-content creators (the deliverable feeds paid ads, not just organic).
- Two-tier program: free = 30% commission, 1 post/story (volume); paid = flat fee (start $300) + free product, 1 dedicated post/reel (no stories — can't run as paid-partnership ads).
- Vetting gate: 2–5% engagement · >20k followers · >80% real followers.
- Test budget: $3,000/mo → 3–5 influencers at $300–$500 each.
Action items — tick when done
🇩🇪 Miranest Germany2
Client MTG💼 client2026-07-09
US ads are highly efficient — cutting budget ¥10M→¥3.6M held 87% of conversions (172→150), proving the old spend was wasteful. The new AI-animation alopecia video (F03/LP1) is a top US performer (15 conv @ ¥7,900 CPU). For Germany: solve ad cannibalization + the 250-ad limit (maxed 236/250 on the shared Miranest Global page) by launching a new third-party FB page. The US influencer report is done — only client (T.) approval blocks shipping + launch.
Decisions
- US spend was inefficient — the lean strategy wins. ¥10M→¥3.6M budget cut retained 87% of conversions (172→150). Won't scale linearly, but the current US approach is proven efficient.
- AI animation is the new US creative model. The educational AI alopecia video (F03, LP1) is a top performer — 15 conversions, ¥7,900 CPU — and avoids the AI "uncanny valley" of fake people. Replicate this style.
- Germany — go around the shared page. Two problems on the shared `Miranest Global` page: (1) cannibalization (our agency + PSJN competing), (2) a 250-ad limit maxed at 236. Fix = launch a new "third-party" FB page (same ad account, pixel, LP — isolate the page variable). An ad-limit increase is unlikely (tied to ~$100k spend).
- Extra DE tests. Niche educational IG accounts (e.g. "Pomeranian Health") to grow organic → repurpose winners as paid "partnership" ads; test German breed keywords ("Zwergspitz" for Pomeranian).
- US influencer program ready. Full report done (followers, comments, commissions, goals); only the client's (T.) approval blocks shipping product + launching campaigns.
Action items — tick when done
Miranest x PSJN — project update🤝 partner2026-06-29
Checkout converting <50% — a visible "Custom Taxes" line on orders >€150 reads as a surprise fee and drives abandonment; hiding/re-labeling it is the top priority. Plus: A/B-test a simpler offer page, €20/post fee to PD's team for influencer sourcing, and benchmark competitor Marmalade.
Decisions
- Hide the customs line at checkout (orders >€150) — highest priority, next few days.
- A/B test a simpler, less text-heavy product-page offer.
- €20/post paid to PD's team for each influencer post they source (starts this month).
- Benchmark Marmalade as the primary competitor for creative + strategy.
Action items — tick when done