Contact

← Back to use cases

Use case #01

EasyEnd: a platform delivered solo, from idea to launch

Market research, design, hosting, AI-assisted development, payments, GDPR, MarTech, campaigns and analytics.

EasyEnd, simple steps before and after a death

Context

Before a death, people need to plan ahead: wishes, end of life, estate. After it, families face wills and dozens of formalities with no clear guide. EasyEnd supports both moments, step by step: guided journey, generated documents, checklist, pre-filled letters, freemium model. An ongoing venture that I founded and run myself, with no team and no agency budget.

What I did, step by step

EasyEnd home page: plan your estate or handle formalities after a death
The home page: two journeys, planning ahead or handling a death.
EasyEnd pricing page: free journey, paid documents, AI assistant
The offers: free journey, paid options, AI assistant.
EasyEnd personal area: progress tracking and generation of the wishes document
The personal area: progress, documents, retention.
EasyEnd article on mobile: how long formalities take after a death
Content, designed mobile first.

1. Market research before the first line of code

Search volumes, competition, sector rules. I found three factual errors in my own acquisition plan, including a compliance one: ad retargeting is banned on bereavement topics by the major ad networks. The strategy was revised before a single euro was spent.

2. Design and communication

Journey, tone, identity, mascot. The topic is sensitive: restraint beats effect. Pages are written to be understood by a grieving person, not to impress.

3. AI-assisted development

Next.js 15, TypeScript, PostgreSQL through Prisma. AI writes much of the code; I decide the architecture, review and correct. Fixes go into the shared component, never screen by screen.

4. Payments and emails

Stripe for paid plans: entitlements, webhooks, idempotency, test/live switch. Transactional email (Resend) and marketing email (Brevo) on shared HTML blocks, with SPF, DKIM and DMARC. Follow-ups automatically exclude paying customers.

5. AI assistant

An assistant grounded in the site content: RAG, tool calling, cost guardrails. After optimization it uses 2 to 3 times fewer tokens per exchange, with the same answer quality.

6. GDPR written into the code

In-house consent banner instead of a subscription, a consent-proof table for anonymous visitors, automated purging based on retention rules, encrypted admin export.

7. Launch and measurement

Hosted in France on a self-managed VPS: Coolify, Docker, Traefik, automatic certificates, zero-downtime deploys for about €9 a month. GA4 and Tag Manager in advanced consent mode, campaigns with landing-page A/B testing.

Pitfalls and decisions

EasyEnd mascot pointing the way
  • An ad platform summary screen can show something other than what is really saved. Publish paused, verify, then activate.
  • The acquisition ceiling is computed from contribution margin, before any spend. It is not a gut call.
  • A keyword-tool export is raw data: I measured a 3.3 to 4.2 times overestimate of real volume.
  • Consent was not replayed at the right moment, so the first page of each visit went unmeasured.

Outcome

A live freemium platform and an ongoing venture: I'm the founder and I run everything myself (product, infrastructure, compliance, acquisition and support). No business figures published here: this case shows execution, not a sales pitch. Want to know more? Get in touch.

See EasyEnd online →

Stack and methods

Next.js 15
TypeScript
PostgreSQL / Prisma
Stripe
Resend
Brevo
Coolify
Docker
Traefik
GA4
Google Tag Manager
RAG
→ Next case: Governance that serves delivery

Facing something similar?

Get in touch