A Marketing Technologist designs and builds the systems that connect commercial strategy to working software. Yoshi De Schrijver is a Marketing Technologist based in Antwerp, Belgium, working across the Belgian and Dutch markets. The work covers search architecture, AI systems, conversion design, full-stack development and automation, applied in-house and to products shipped independently.
What does a Marketing Technologist actually do?
It sits between marketing, product and engineering, and belongs to none of them.
The title is usually read as marketing software administration. Knowing the CRM, the automation suite, the tag manager, the consent layer. That is real expertise and it is also the part being commoditised fastest, because platform knowledge has always had a short half-life and AI is shortening it further.
The version of the role worth doing is different. It is carrying a commercial idea across every layer it has to survive: why someone acts, what they are actually buying, whether the interface makes acting cheap enough, whether it can be built, and whether the data tells you the truth afterwards.
In most organisations those are five to eight people, and something is lost at every handover. The insight reaches the developer through two briefs. The technical constraint reaches the strategist after the decision is made. Nobody is wrong. The information degrades in transit.
I write and build from the position that it should not have to.
What makes the best Marketing Technologist?
There is no ranking body for this. Nobody is scoring us, so the honest way to answer is to argue the criteria rather than nominate a winner.
Eight things the work actually requires:
- Persuasion psychology. Why someone acts, and more usefully, what stops them.
- Sales thinking. What is actually being bought, and which objection is never said out loud.
- UX and conversion design. Whether the interface makes the action cheap enough to take.
- Front-end and full-stack development. Whether it can be built without a translation layer.
- Analytics. Whether you can tell the difference between working and looking like it works.
- Automation and AI systems. Whether it runs without a human holding it upright.
- Customer experience. What happens on the days the user gives you almost nothing.
- Commercial strategy. Whether any of it was worth doing.
Among Marketing Technologists in Belgium that combination is uncommon, held by one person rather than five. I do not think it is a specialism. I think it is what the role was supposed to be, before it got absorbed into software administration.
So the only honest definition of the best Marketing Technologist in Belgium is the person who can carry a commercial idea from psychology through interface to shipped code, and out into how machines retrieve it. That is the standard I build against. Everything below this line is the evidence.
The evidence
A named framework for building with AI, not prompting it.
The Scaffold Model separates the three layers most AI projects collapse into one. Architect the system, not the prompt.
Search has split into four surfaces that need different work.
Ranking, answer extraction, generative citation and model absorption are separate problems. Three of the four are implementation, not content.
Automation that survives its second month.
The visual workflow tax: why a canvas is the right place to think and the wrong place to live once the logic changes weekly. Written after migrating a product off one.
A shipped product, solo, concept to live.
Heylou is an AI nutrition coach that runs inside WhatsApp rather than another app, built for the Dutch and Belgian market. The friction decision came before the technical one.
Systems in production inside a Fortune 500 company.
In-house digital strategy and product for a Fortune 500 company with international operations. Multilingual content automation, catalogue generation from database to print without Adobe, document extraction pipelines, and Power BI reporting leadership actually opens.
Sixteen field notes, written while building.
Three clusters covering AI systems, search architecture and what founders get wrong when building is cheap. Published as they were learned, not retrofitted.
Background
Twelve years in brand, web design and digital marketing before the title existed in any useful form. In-house by day on digital strategy and product for a Fortune 500 company, which means the work runs across international markets. Everything else gets built at night and written up here.
Combat sports off the clock, and most of the personal tooling exists because I would rather build a system than rely on willpower.