Guides
What we learned by doing it.
One guide per skill: the method, the pitfalls, and the projects where we applied it.
- Project management
How to scope a digital project, from idea to launch
A project is scoped in three stages: test the idea before building, ship a short first version, then measure to decide what comes next.
Read the guide - Web app
Building a web app: the steps, the technical choices, the pitfalls
A successful web app starts small: one complete journey, a proven stack, and a production deployment in the first week.
Read the guide - AI and agents
AI agents: connecting them to real data
An AI agent is worth the data and tools you give it, and the guardrails you put around it: cited sources, review, a capped budget.
Read the guide - E-commerce
Launching a Shopify store: the method, from theme to first order
A Shopify store launches in five workstreams: the brand, the catalogue, the theme, the checkout and the measurement. In that order.
Read the guide - E-commerce tracking
E‑commerce tracking: GA4, Search Console and Merchant Center
Three free tools are enough to measure a store from first visit to order: GA4 for the journey, Search Console for Google, Merchant Center for products.
Read the guide - Google Ads
Google Ads for an online store: where to start
Google Ads amplifies a store that already converts. Work out your margin first, make purchase tracking reliable, then start small on products.
Read the guide - WordPress
WordPress and WooCommerce: delivering a turnkey website
WordPress is still the right choice for a site the client runs on their own. It lasts with few plugins, regular updates and tested backups.
Read the guide - Mobile apps
Building an iOS and Android mobile app: from mockup to stores
A single React Native codebase covers iOS and Android. The real work is elsewhere: the journey, the API, payment and store review.
Read the guide - DevOps and hosting
Hosting a site or an app on a VPS: the DevOps method
A well-run VPS hosts a site, an API and a database for a fixed cost. It takes four disciplines: secure, automate deployment, back up, monitor.
Read the guide - Multilingual SEO
Multilingual SEO: hreflang, translated URLs and mistakes to avoid
Each language has its own URLs, linked together by reciprocal hreflang tags. Translation also covers keywords, titles and the URLs themselves.
Read the guide - AI agency
AI digital agency: what it does, and how to choose one
An AI digital agency designs websites, apps and stores, and connects agents and automations to them that work on your real data.
Read the guide - Nomad agency
Nomad agency: working with a fully remote digital agency
A nomad agency has no offices: the team works remotely, from several countries. For the client, the method matters more than the address.
Read the guide - Remote work
Running a remote digital project: the asynchronous method
A remote project succeeds with three habits: write decisions down, deliver every week on an address the client opens alone, and keep a few shared hours.
Read the guide - AI in business
Bringing AI into a small business: where to start
You bring AI in through a specific, repetitive, checkable task, not through a tool. The first project should be measured in hours saved after one month.
Read the guide - AI audit
AI audit: finding the use cases that pay off in your business
An AI audit lists the tasks that can be automated, scores each on gain and feasibility, and comes out with three projects to launch in order.
Read the guide - AI automation
AI automation: which tasks to hand to an agent
Classic automation follows rules. AI automation handles what requires reading and judging: sorting, summarising, drafting, reconciling.
Read the guide - Automation
n8n, Make or custom code: which choice for AI automation
A no-code platform is enough to connect tools with a few AI steps. Custom code wins when volume, logic or sensitive data go beyond that frame.
Read the guide - AI chatbot
AI chatbot for your website: getting a customer service assistant right
A good AI chatbot answers from your content, cites its source, can say it does not know and hands over to a human at the right moment.
Read the guide - RAG
RAG: getting an AI to answer from your documents
RAG makes the AI look up the useful passages in your documents before answering. It then answers on your content, with the source, with no retraining.
Read the guide - MCP
MCP server: connecting Claude and ChatGPT to your tools
An MCP server exposes your tools and data to an AI assistant in a standard way. The assistant can then read and act in your systems, with the permissions you set.
Read the guide - AI and e-commerce
AI for e‑commerce: the use cases that pay off
In e-commerce, AI pays off first on the catalogue and on management: product pages, translations, spotting stock-outs and loss-making ads. The chatbot comes later.
Read the guide - AI budget
How much an AI project costs: what drives the price
An AI project has two costs: the build, which mostly depends on the state of your data, and the monthly running cost, which depends on volume and can be capped.
Read the guide - AI and compliance
AI, GDPR and the AI Act: what a business should check
Three questions cover most of it: which data goes to the model, do people know they are talking to an AI, and who reviews what it produces.
Read the guide - AI search
Generative engine optimization (GEO): getting cited by ChatGPT, Gemini and Perplexity
To be cited by AI engines, a site must be readable by their crawlers, answer questions clearly and be recognised elsewhere on the web.
Read the guide - Content and AI
SEO content with AI: producing without losing rankings
Google does not penalise AI, it penalises content with no value. AI speeds up the writing; experience, facts and review stay human.
Read the guide - AI development
Building a custom AI application: from idea to production
An AI application is an application first: accounts, data, interface. The model is only one component, which must be testable, capped and replaceable.
Read the guide - MVP
Building an MVP with AI: testing your idea in a few weeks
AI shortens the making of an MVP, not the thinking. The scope remains one complete journey, put in front of real users as early as possible.
Read the guide - Choosing a provider
AI agency, freelancer or in‑house team: how to choose
A freelancer for a specific component, an agency for a complete product to deliver, an in-house team when AI becomes the lasting core of the business.
Read the guide - Project brief
AI project brief: the one‑page template
A good AI brief fits on one page: the task, the data, examples of good answers, what the agent is not allowed to do, and the success measure.
Read the guide - Development and AI
AI‑assisted development: what it changes for your project
Coding agents write fast. A project still holds together through what they do not do alone: scoping, architecture, review and tests.
Read the guide