Video summary

Milia Matcha : Elle perd des millions tous les mois

Main summary

Key takeaways

Business

Business case overview (what happened + why it matters)

  • Milia Matcha (France/Japan) launched a matcha brand in 2023, backed by a YouTuber/founder whose audience created early demand.
  • Early traction (launch economics)
    • Flagship product sold out in ~2 hours (inventory for ~1 year).
    • After 6 weeks: ~€500k revenue
    • Profitable by month ~1.5 (before Christmas), despite frequent out-of-stock periods.
    • Forecasts / targets mentioned
      • 2024: €3M revenue
      • 2025: €5M revenue
      • “5M by 2025 starting from scratch”
    • Ops footprint expanding
      • Offices in France and Japan
      • Shipping worldwide
      • Opening warehouses in Germany, Spain, UK; preparing for Asia

Core thesis: differentiation is strong, but growth is “mathematically ceilinged” by founder-driven demand. Paid ads currently cover only part of the funnel.

Key metrics & KPIs cited

Revenue

  • €500,000 at 6 weeks post-launch
  • €3M (2024) target / projection
  • €5M (2025) target / projection
  • “5M over the last 12 months” (brand performance claim)

Traffic / acquisition signals

  • 98,000 visitors in a month, supported by high founder YouTube presence + promo codes
  • 30,000 visitors in the following month with less content/presence
  • ~50,000+ “routines” completed in 2 months (subtitle: “500,000 routines”)

Reputation / product quality

  • 4.6/5 on Trustpilot, ~1000 reviews

Team / structure

  • 9 permanent employees
  • Distribution ops (warehouses) planned/active: DE/ES/UK

Ad volumes (current state)

  • 45 active ads (MTA/marketplace-type platform)
  • 65 on Google
  • 650+ on TikTok
  • The issue is framed as creative messaging + funnel coverage, not spend volume alone.

Main business strategy & playbooks/frameworks referenced (implicit “funnel machine” model)

1) Mature-market positioning logic (“difference + remove the reason not to buy”)

In a mature, crowded matcha market, success is framed as:

  • Eliminate the #1 purchase objection
  • Differentiate vs. the crowd so a consumer has a reason to choose you

Applied to:

  • taste
  • visual identity (see below)

2) Funnel model (“black onion” top/middle/bottom levels)

Acquisition problem described as:

  • Top-of-funnel (TOFU): people who don’t know matcha/brand yet → education/interest creation
  • Middle-of-funnel (MOFU): interested but hesitant → proof of differentiation (why this brand)
  • Bottom-of-funnel (BOFU): ready-to-buy → offers, reviews, purchase triggers

Current gap: ad strategy is mostly BOFU-only, while the founder’s content supplies TOFU + MOFU.

3) Acquisition machine playbook (continuous creative testing + stage-specific landing pages)

Proposed fix: build an “acquisition machine” with:

  • Continuous creative refresh
  • Multiple creatives per funnel stage
  • Dedicated landing pages per ad angle/flavor/topic

Concrete differentiation decisions (product + brand) and how they drive demand

Decision A — Taste strategy: solve matcha bitterness via flavoring

  • Traditional matcha is described as bitter/vegetal.
  • Milia Matcha is positioned as “matcha without bitterness” via flavors (e.g., strawberry, vanilla, banana).
  • Market impact claim
    • Product line launched with 3 SKUs
    • Vanilla sold out first and became the bestseller
    • Response indicates demand creation
  • Addressable market expansion
    • Not just matcha lovers—also people who tried matcha and didn’t like it
    • The logic: market size is multiplied, not merely improved

Decision B — Visual identity strategy: pink “break the rules”

  • Competitors use green as the default visual code → blends in.
  • Milia Matcha uses pink cans/universe for instant recognition:
    • “Recognize it in a quarter of a second” among green brands
  • Framed as both aesthetic + strategic:
    • Strong recognition in social feeds and in-store contexts

Operational/launch execution points highlighted

  • Japanese supply collaboration is complex (language/culture barriers).
  • Tailor-made products were developed with Japanese producers.
  • Packaging engineered for industrial constraints.
  • Launch timing delays occurred.
  • Early development investment included ~€2,500 “even without knowing if it would work” (early experimentation cost).
  • Key emphasis: launching a brand requires months of invisible operations before sales—often underestimated.

“Where money is being left on the table” (structural growth limits)

Diagnosis: founders’ content is doing 2/3 of the funnel work

  • When founder content spikes → 98k visitors
  • When content/presence drops → 30k visitors
  • Ads don’t fully replace that; advertising contributes ~1/3 of needed funnel function, while the founder provides the rest.

Why paid ads underperform (despite volume)

Current ad creatives described as:

  • 6–15 seconds
  • beautifully filmed
  • no voice-over
  • no hook / no argument
  • no explicit problem-solution

Result: they convert people who already know the brand, but don’t create demand.

Three structural constraints named

  1. International scaling

    • In new countries, the audience hasn’t been “educated” via founder content.
    • Expansion (e.g., DE/ES/UK) doesn’t automatically bring the TOFU engine.

  2. Valuation / sellability

    • Founder-led brands can be discounted for M&A because growth depends on a single person.

  3. Mathematical ceiling

    • Brand growth is capped by one person’s audience, while matcha demand is larger.

Actionable plan (if the speaker “took the reins”): acquisition machine rebuild

Project 1 — Transform ads into a full-funnel machine

Key rule: acquisition doesn’t run on “45 active ads”; it runs on:

  • dozens of new creatives tested continuously every week

Reason:

  • creative wears out in cold traffic in a few weeks → if not refreshed, CPA rises and budgets get cut.

Creative diversification (new formats)

  • UGC with real customers (facecam), with “why I chose this matcha”
  • Spoken-hook formats (actors or on-camera hosts)
  • Comparisons / before-after recipes
  • Objection-led opens, e.g.:
    • “I’ve always hated the taste of matcha—try this one”

Funnel coverage build (“black onion” levels)

  • TOFU creatives
    • matcha benefits
    • switching from coffee (energy without crash; morning routine)
  • MOFU creatives
    • educate differentiation:
      • why taste differs
      • why it’s flavored
      • Japanese origin / producer credibility (as relevant)
  • BOFU
    • offers, reviews, new products (what they already do)

Landing page architecture

  • Don’t send everything to one product page.
  • Create dedicated product pages per aroma/flavor/angle:
    • matching the ad message improves conversion after the click

Project 2 — Build an influencer/affiliate engine to remove founder dependency

Current limitation: only one influencer is actively representing the brand (the founder), even though she could activate many creators.

Proposed channel: multi-creator performance affiliate system (especially suitable for TikTok Shop)

How it would work

  • Recruit creators who speak to:
    • matcha
    • coffee/energy
    • morning routines
    • well-being
  • Range: nano → big accounts
  • Provide:
    • product samples
    • real incentives to post
  • Commission on sales via:
    • dedicated tracked creator pages
    • gamification + tiers + rewards based on results
  • Risk control: “creator only wins if they sell” → near-zero risk for the brand

Expected compounding effect (“snowball”)

  • When 10–20–50 creators post in the same period:
    • product appears across multiple homes
    • purchase probability rises
  • Brand shifts from “single-person owned” to system-driven.

Growth timeline / targets (execution roadmap)

  • Next 6 months: become France benchmark for matcha
  • Next 12–24 months: lead Europe market
  • 3–5 years: no structural reason it can’t scale globally

Reasoning:

  • once the acquisition machine works, it can be duplicated, translated/adapted, and supported with local ambassador relaunches.

Key business leadership/management insights stated

  • Don’t assume influencer leverage is enough: community helps, but it’s not a guarantee (competitor examples mentioned where influencer brands failed long-term).
  • The true differentiator for scaling is the “machine,” not just product quality
    • “A good product doesn’t sell itself”
    • “A good brand doesn’t sell itself”
    • Companies that scale beyond others build systems around demand generation, not just product.

Concrete examples/case references used to support the thesis

Competitor/influencer brand failures referenced as cautionary tales:

  • Cyprien’s brand disappeared
  • MXM repositioned Spacefox from clothing to jewelry
  • Vot Caprod’s “Unicorne” brand liquidation (Jan 2024)
  • Valousose burgers liquidation (referenced similarly)

Advertising mechanics example:

  • Ads work when they convert existing awareness, but fail to generate initial demand.

Presenters / sources

  • Sofian (e-commerce founder/coach; presenter/expert in the video)
    • Background: 8 years e-commerce, multiple French brands
    • Stay Focus coaching program (team of five experts for 4 years)

Original video