Structured niche-testing for Meta - and Google when data says so.
01 / Where You Are Right Now
You already have PMF in one niche. Neurodivergent parents found Sharp, it works, and they're converting on paid Meta after two weeks. Replicate it across four or five more segments.
AppsFlyer is your source of truth - but you're not confident it's correct. Two weeks of campaigns, trial still open, attribution discrepancies with Meta. Week one: audit, test event firing, confirm data is trustworthy. No scaling on numbers nobody trusts.
€5K needs a testing matrix, not vibes. Without a framework it's noise. With structured niche isolation and weekly cut/scale decisions it's a system.
Month one is for learning, not ROAS. You have a €200 CAC ceiling and a 30-day trial window that hasn't closed yet. Before scaling to €15K+ we need to know which niches actually hit that number - and which don't. Month one gives us that answer.
02 / Niche Testing Strategy
You came in with sharp hypotheses. Here's how we sequence them into testable campaigns.
Already converting. Tight-knit community, fast word-of-mouth once a product helps. We expand into sub-segments (ADHD vs. dyslexia vs. dyscalculia) and test if specificity improves CPT.
Why we have conviction here: ABA therapy for kids with ADHD, ASD, learning differences in the US. Parents converted at €80-120 per consultation. Sharp removes all access barriers. Same psychology, lower CAC ceiling.
Maghreb, West Africa, Turkey, South/East Asian diaspora. Language barrier = can't help with homework. Sharp may be the only tutoring that actually works for them. Diaspora communities share recommendations fast.
Outside Paris, Marseille, Lyon - tutors are scarce or don't exist. "Quality tutoring, wherever you are" is an advantage metro-focused competitors can't claim. Clean geo isolation in testing.
Studying in French but it isn't the dominant language. Academic French lags. Well-educated, investable parents. Strong presence in Paris 16e, international school zones, Lyon.
Homework falls through between two homes. Sharp is always available - no parental coordination required. Angle: stability for the kid, not blame toward either parent.
03 / Channel Strategy
One network, full focus. Prove niche-to-trial on Meta first. €5K is the right size for one channel done properly.
Right channel for parent targeting in France. Behaviour layering isolates each niche cleanly. Free trial = near-zero friction. Funnel works when creative speaks to the right parent.
Highest intent possible. "Soutien scolaire TDAH Paris" = parent already in market. Added once Meta gives clear cost-per-trial benchmarks to model against.
03b / Trial Funnel
Generic "AI tutoring app" ads don't convert. Lead with the specific problem - ADHD, rural isolation, language barrier. Parent reads line one and thinks "this is for us."
Retarget non-converts: 30 days, no commitment, cancel anytime. Map in-app events through AppsFlyer to find which early actions predict paid conversion.
At day 30, trial-to-paid data comes in. Identify the event that best predicts conversion, shift retargeting toward users who hit it. Light-touch: progress reminder, what paid unlocks. Never cold.
04 / Deliverables
Tracking review in week one. Structured testing table, not guesswork. Every decision explained in writing, every week.
05 / Add-Ons
Each add-on activated when unit economics make it the logical next step - not before.
Google Ads
High-intent search for tutoring in French. Meta angles transfer directly - no restart from zero. Activate once Meta proves which segments convert.
TikTok Ads
You're already running TikTok organic. Paid TikTok amplifies what's resonating - useful for the kid-installs-convinces-parent funnel.
AI UGC Video Ads
AI-generated UGC-style parent testimonials for Meta. €50 per video. Fast, cheap, effective for rapid niche-specific testing without creator lead times.
SEO + GEO (AI Search Visibility)
Rank in ChatGPT and Gemini for "meilleure app soutien scolaire France". Press release across 300+ outlets structured for AI indexing - our Fintech SaaS client went #1 in Gemini within two days.
06 / What to Expect
Week 1
AppsFlyer audit: test events, flag attribution issues, recommend what to track. Audience research and ICP mapping per niche.
Week 1–2
Matrix finalised. Niche audiences structured with messaging angles. French creative briefs delivered. Campaigns built in Meta and ready.
Week 2–4
Live. CTR, CPC, install rate, trial starts tracked per niche. Early signals drive budget. Weekly call + written update every Monday.
Month 2–3
Enough data to draw conclusions: which niches convert, at what CPT, which angles scale. Budget moves from €5K to €15K on solid ground.
07 / Investment
Flat management fee per channel. No hidden costs. Creatives separate.
Meta Only - Recommended Start
per month
Meta + Google Ads
per month - $300 off vs separate retainers
08 / Our Work
Mobile app + AppsFlyer. Parent niche with learning needs.
Same AppsFlyer multi-event methodology. Don't optimise for install - optimise for the event that predicts retention.
Mobile app, multiple in-app steps. AppsFlyer showed users completing three specific actions within 48 hours retained far better. Shifted to that event - CPI dropped, ROAS up 37%.
For Sharp: find the in-app action that predicts paid conversion, shift optimisation there, CAC becomes predictable.
Same parent persona. Same urgency dynamic. When the message matches the pain precisely, CAC is manageable even in a small niche audience.
ABA therapy and specialist learning for kids with ADHD, ASD, learning differences in the US. Goal: consultations and trial sessions. High urgency, tight-knit parent communities.
Sharp's unit economics are better - no per-session limit, subscribers retain for months. CPT under €50 is realistic for this segment in France.
Ready when you are
Week one: validate the data. Week two: campaigns live. Every decision after driven by what the numbers say.
Telegram
@svetagreenOney Studio / Performance Marketing — Sharp (BendingWords) / Gustav