I run a $10M education business end to end.Dad.Book reader.Terrible with gardening.

Diagnosis: $10M catalogue, zero recurring base

The education business does ~$10M selling courses. Freya alone: 17 courses, strong one-time sales, and every month starting from zero. Students finished and disappeared; nothing captured ongoing value. The job was to add a recurring layer without killing the catalogue that already worked.

Strategy: price, package, position

Buyer research shaped the ladder: annual anchor at $299/year (full catalogue + future releases), with monthly and quarterly alternatives. Positioning as creative identity, not “access to courses.” Product = academy membership; Place = Studio + checkout; Promotion = identity-led messaging. Objective: ARR and plan mix, not a vanity launch. Today 51% of subscribers sit on annual, the highest-LTV tier.

Execution: Freya Studio as the product

Member product behind the subscription: courses, 450+ brushes, community, summit access. Scoped what the sub includes and excludes so it sits beside standalone sales without cannibalising them. Migration and sign-in for students who bought before the platform existed.

Execution: checkout, email, analytics

FunnelKit checkout with order bumps and plan-switching. Klaviyo: pre-purchase nurture, post-purchase onboarding, renewal. First clear subscription analytics layer: MRR, ARR, plan mix, churn signals, so retention had numbers on day one.

• Product definition: Scoped what the subscription includes (17 courses, 450+ brushes, all future releases), what it excludes, and how it sits alongside standalone course sales without cannibalising them
• Pricing architecture: Set $299/year as the anchor with monthly and quarterly alternatives, based on research output, not guesswork
• Landing page: Built and copywritten from scratch. Conversion-optimised structure: lead with identity, follow with proof, close with value math
• Checkout flows: Implemented FunnelKit checkout with order bumps and plan-switching logic
• Email sequences: Built pre-purchase nurture, post-purchase onboarding, and subscription renewal flows in Klaviyo
• Analytics layer: Set up full subscription tracking, MRR, ARR, plan mix, churn signals, so the business could see recurring revenue clearly for the first time

Optimisation: the control surface

Snapshot Apr 28, 2026: $18,345 MRR (+5.2% MoM) → $220,144 ARR run rate · ~1,384 paid subscribers. Same month’s MRR×12 is the ARR figure: one dashboard, not two stories.

Optimisation: what the numbers pushed next

Recurring base that did not exist before this project. 51% annual is the retention bet; next loop is expanding lower entry-point tiers in 2026 without wrecking the $299 anchor. Diagnosis starts again from the dashboard.

• $18,345 MRR with +5.2% month-on-month growth trend
• $220,144 ARR, recurring revenue base that didn't exist before this project
• ~1,384 active paid subscribers across three plan tiers
• 51% of subscribers on the annual plan, the highest-LTV, lowest-churn tier

Explore

See the live academy offer.

The category was full of webinars, empty of events

We looked at digital art education events the way a buyer would. What people wanted: working-artist workflows, not polished tutorial streams. What the market shipped: passive webinars with no community layer. The gap we chose to fill: interactive, multi-speaker summits with real audience connection, not another free Zoom.

Target the Procreate community, not “creatives”

Audience narrowed to Procreate and digital illustration, not a vague creative TAM. Positioning: premier digital art summit, not webinar. Objectives: registrations, attendance, repeat editions. Offer choice: 30+ speakers per event as value a thin webinar cannot fake. Price and urgency served those objectives.

Checkout that holds conversion

Mobile-first checkout: offer stack above the fold, trust next to pay, no account wall. Headline, speaker order, and price anchoring were tested until checkout conversion held above 60%. Packaging and price as execution, not a slide.

Post-purchase engagement

VIP access with schedule, community links, and WhatsApp onboarding. Retention starts the minute they pay: zero confusion between purchase and showtime.

Tracking while the campaign is live

Summit admin report: pacing, projected tickets and revenue, AOV, funnel glance. One source of truth for media buyers and ops before doors open; budget followed the dashboard mid-flight.

How 46K registrations actually arrived

Email sequences drove ~70% of registrations. WhatsApp automation held 85%+ open rates. Landing and paid filled the rest. Acquisition was a channel system, not a single ad set, measured across editions totaling 46K registrations, 19,274 hours watched, $300K+ revenue.

• Landing pages: 60%+ checkout conversion
• Email sequences: drove 70% of all registrations
• WhatsApp automation: 85%+ open rates
• Community layer: turned attendees into advocates
• Speaker recruitment: 30+ international artists per event
• Reach: 15M+ impressions across channels

Optimise across editions

Four summit editions. What repeated (speaker-grid LP, email-led volume, WhatsApp ops) stayed in the playbook. What underperformed got cut before the next launch. Next diagnosis starts from those numbers.

• 46,000 registrations
• $300K+ revenue
• 19,274 hours watched
• 180+ countries
• 8.8/10 average attendee rating

Diagnosis: vendors were the growth tax

Summit, courses, and community lived in separate tools. Checkout, access, and content could not share one student identity. This case is not another campaign; it is the constraint under Summit and Freya: every vendor handoff slowed ops and broke entitlements.

Strategy: one member stack, two brand spaces

Twotap Studio as the operating layer: courses, community, summit access, checkout entitlements. Objectives: one identity, one commerce path, reusable for Twotap Access and Freya’s Academy space. Refused to run the next summit on a stitched five-vendor stack.

Execution: migration without stranding buyers

Sign-in and migration for students who bought before Studio. Retention = not stranding legacy buyers. One identity across courses, summit, and community.

Execution: what the platform actually runs

Member access, course library, community + in-app summit surface, commerce bridge. Live at twotap.art and studio.twotap.art: courses, community, and checkout no longer stitched across vendors.

• Member access: sign-in, migration for prior purchases, one identity across products
• Courses & library: structured learning with brushes and resources in-product
• Community & summit: feed and in-app summit surface without a second tool
• Commerce bridge: storefront and checkout at twotap.art feeding Studio entitlements
• Multi-space: same platform powers Twotap Access and Freya's Academy space

Optimisation: the stack behind the proof

Same foundation under Digital Art Summit editions and Freya Studio. When a new product needs a surface, we extend Studio; we do not add another vendor. Metric of success: one member stack, two brand spaces, full course · community · checkout path.

• One platform behind summit + academy products
• Twotap and Freya spaces on the same Studio foundation
• Courses, community, and checkout no longer stitched across vendors
• Live at twotap.art and studio.twotap.art

Explore

See the live store and Studio.

Diagnosis: agencies stuck in spreadsheets

In Slovenia, agencies ran influencer campaigns in spreadsheets: no discovery, no audience verification, no content tracking. Global tools were built for other markets; nothing local owned the ops job. Avalan’s brief: the operating system agencies actually need, not another marketplace pitch.

Strategy: nine agencies, yearly contracts

Target: the nine largest marketing agencies in the country, not freelancers, not brands first. Positioning: multi-campaign ops. GTM: founder-led outreach and yearly contracts for lower churn. Sell before a paid acquisition budget existed.

Execution: brand, decks, closed meetings

Distribution through founder-led sales. Brand system and pitch decks for a zero-ad launch. Acquisition = meetings closed, not clicks bought. Signed 6 of those 9 top agencies.

• Product: influencer discovery across 80M+ profiles, audience authenticity analysis, automated content tracking
• Sales: signed 6 of 9 biggest marketing agencies in Slovenia to yearly contracts
• Clients: L'Occitane, Philips, Publicis Groupe, Baby Center, Dedoles, Pristop
• Brand: designed entire identity from scratch
• Pitch decks: wrote and delivered all enterprise sales presentations

Optimisation: month-one ARR

€15K ARR in month one: paying agencies before a marketing budget existed. Proof that the agency-first bet worked; roadmap and sales messaging followed who actually paid.

• €15K ARR in month one
• 690+ campaigns tracked
• 18,928 verified posts
• 6 of 9 top Slovenian marketing agencies as clients
• 80M+ profiles in discovery database

Explore

See the live platform.

Diagnosis: Freya’s price is invisible at 11pm

Operator read from Freya demand: $299/year converts, and is invisible to someone scrolling TikTok at night. No mobile-native path offered real artist feedback at an impulse price. Segment we aimed at: people who will pay coffee-money to start tonight, not commit to an academy.

Strategy: impulse product, App Store GTM

Positioning: personalised drawing coach in your pocket. Price: $4.99/month. Offer shape: onboarding quiz, 15-minute lessons, human feedback aimed at a 48-hour turnaround. Objective for this case: prove a growth operator can ship native distribution, not run paid UA yet.

Execution: product + ops without a team

Four-tab architecture, Bunny video, Supabase auth, Klaviyo lifecycle, Slack ops alerts. Built with AI-assisted coding; zero developers hired. Tracking and notification hooks in from day one.

• Product: Specced the full four-tab architecture, onboarding quiz, botanical streak system, lesson structure, and feedback submission flow
• Build: Built entirely with Claude Code, no developer hired. Integrated Bunny (library 16137) for native video delivery and Supabase for database and auth
• App Store: Cleared three Apple rejections on submission requirements. Learned more about App Store review policy in 25 days than most developers learn in a year
• Design: Full visual identity and UI direction. All graphics commissioned separately, everything else built solo
• Infrastructure: Supabase edge functions for email triggers, Klaviyo for lifecycle flows, Slack webhooks for submission notifications and ops alerts
• Timeline: 25 days from first line of code to live on the App Store

Optimisation: honest early stage

No paid acquisition yet. No vanity install story. The case study is the ship: idea to App Store in 25 days. Next loop starts when usage shows which lesson paths and feedback turnaround to double down on.

• Live on the App Store, May 2026
• Built and shipped solo in 25 days with no development background
• Three Apple rejections cleared and resolved, shipped anyway
• No paid acquisition yet. The proof of concept is the product existing.

Explore

Download on the App Store.

Diagnosis: skill without a repeatable system

Fifteen years of photography. One-off charity shoots do not scale impact. Intent: put €500+ quality work toward people who need it. Gap: no repeatable system where sessions fund charity 100%.

Strategy: a year-long alpha, then open the model

Positioning: good work that does good. Identity designed for Worthly, not borrowed from a studio portfolio. Experiment: once a month for a year, session quality at €500+, donation 100% to Botrstvo and Projekt Vida. Alpha was me showing up; the platform is the next step so other pros can run the same loop.

Execution: people & proof

Proof-first profiles, booking flow, campaign impact. Trust for the next session is the retention loop.

Execution: twelve months of alpha

Twelve months of sessions completed under the alpha rules, brand system built, platform at worthly.org. Honest status: the year of showing up is done; opening the model to other professionals is in progress.

• 12 months of sessions completed
• Full brand identity designed from scratch
• Platform live at worthly.org
• Photography workflow optimized for volunteer efficiency

Optimisation: from alpha to platform

Iterate the platform from what the alpha proved, not from a pitch about scale that has not happened yet. Next diagnosis: which pros join, and whether the 100%-to-charity rule still holds when it is not only me.

• 12 months of proof
• 100% to charity, every session, every time
• Now building Worthly into a platform
• Any professional can give their best work, 100% to charity

Explore

Visit Worthly.

Diagnosis: household money is awkward

Built from living it: shared household money is awkward. Spreadsheets lie, banks are cold, couple apps cosplay as finance products. The real job: know why the money disappeared, without turning the relationship into a board meeting. This is a personal ops tool, not a fintech launch story.

Strategy: private PWA, not a pretend bank

For two adults (Domači računi): track, settle fairly, roast the month, keep goals visible. Visual system: forest green, parchment, zero corporate cosplay. No paid GTM: product for the people who live in it.

Execution: what actually shipped

Dual-person balance and settle, expenses and recurring bills, analytics with monthly roast, AI advisor + receipt scan (Gemini), household goals. Next.js PWA with Supabase auth, shipped for real daily use, not a pitch deck.

• Dual-person balance, fairness score, and settle flow
• Expenses, categories, recurring bills, and household goals
• Analytics with charts, calendar, and monthly roast
• AI advisor + receipt scan (Gemini)
• Shipped as Next.js PWA with Supabase auth

Optimisation: two adults, daily use

Live daily for Damjan and Nina. What gets used next month stays; what feels like a bank gets cut. Proof: consumer tools can be sharp and human for a household of two.

• In daily use for a real household
• Full settle + analytics + advisor loop shipped
• Brand and UX that refuse corporate finance tone
• Proof: consumer tools can be sharp and human