Open to senior roles in content and strategy Leo Smits · Amsterdam, NL

AI content systems · Decision design · Creative strategy

I make complicated things clear, compelling and actionable.

I’m Leo Smits. I’ve spent twenty years in communication: from advertising and creative direction to UX, AI and content systems. I work where ideas, language, experience and systems meet.

What interests me most is not simply improving what already exists, but understanding what should exist in the first place. That can mean finding a better sentence, redesigning a journey or questioning the whole idea before anyone starts making it.

Typical assignmentsAI content governance and operating models  ·  Terminology and source-of-truth systems  ·  Decision design for complex journeys

30M+ members  ·  21 markets  ·  Aviation, banking, e-commerce and airport operations

See my work Let’s talk
Leo Smits

Approach

I like to get involved before the solution is fixed.

A lot of communication work starts too late. The proposition has been decided, the journey is drawn and the brief has become a request for words. I prefer to come in earlier, while there’s still room to ask what we’re actually trying to change, what someone needs to understand and why they should care.

The disciplines have changed over the years. The instinct hasn’t: make it understandable and relevant, and make sure it leads somewhere.

What I do

Three ways in. One instinct.

Content systems & AI

Content scales when the thinking behind it becomes reusable.

AI content operating models and governance: prompt architecture, structured source knowledge, evaluation criteria and human review. Underneath that, the foundations AI depends on: terminology, taxonomies, source-of-truth content, localisation, templates, and design systems that keep AI-generated copy and imagery on-brand.

What interests me about AI is what happens once it becomes part of the actual creative and content system.

Content experience & decision design

People rarely need more content. They need to know what matters, what their options are and what happens next.

Journeys, interfaces, CRM, onboarding, loyalty experiences and complex service flows, with particular attention to the moments where people have to interpret information, make a choice or take action. I review those moments with my own framework: Understand → Decide → Act.

Ideas & creative direction

Before the copy deck or the wireframe, I look for the thought that holds everything together.

Propositions, campaign concepts, messaging and activation, and increasingly the way copy, image, AI and interaction work together as one coherent idea. Good creative work starts with knowing what you want people to feel, understand or do. The format follows from there.

Structure is not the opposite of creativity. Good structure gives creativity somewhere to go.

Selected work

Six problems that didn’t fit one discipline

Most of my recent work is for Flying Blue, the loyalty programme of Air France and KLM. Product and marketing overlap there more and more: a journey can be functional and persuasive at once, and an email is both communication and experience. That overlap is where I’m most useful. Before Flying Blue: airport operations, banking and e-commerce.

Making AI drafting faster without losing control

Flying Blue · AI content operating model

When teams started drafting with AI, speed was easy to get. Consistent terminology, verified claims and compliance were much harder.

I designed the operating model around it: prompt architecture, structured source knowledge, evaluation criteria, governance and human review. The AI writes the drafts; the system decides what’s allowed.

Finding out what still needed to be decided

Flying Blue × Open Banking · Mastercard

This project brought together Flying Blue, Mastercard, an open-banking provider, legal stakeholders and product teams. Everyone was working on the same journey, each with a different interpretation of it.

I turned those ambiguities into 17 explicit open questions, made the blockers visible and structured the answers, so every party responded to the same reality.

A useful reminder that content design is often about discovering what hasn’t been decided yet.

Result

Faster alignment between four parties, and terms and conditions that left less room for reinterpretation.

One word for one thing

Flying Blue · Air France · KLM · Terminology and governance

Large organisations often have several words for the same thing, and the same word for several things. It starts as a language problem and quickly becomes a product, brand and operational one.

I built a multilingual glossary and taxonomy with canonical terms, definitions and usage rules. It’s now the single source of truth for websites, localisation and AI systems.

Result

248 terms in English, French and Dutch, shared by two airline brands. That meant less rework and safer translation.

Turning two decisions into one journey

CarNext · E-commerce in 21 European markets

Buying a car online and trading in your current one used to be two separate decisions, each with its own doubts. I wrote the content for a journey that joined them: the car’s condition in three questions, a valuation by email within minutes, and offers with the trade-in value already deducted.

The same years covered the other end of the work: social campaigns with brand ambassador Max Verstappen, including The Fastest Egg Hunt, an Easter game on CarNext’s site with a VIP weekend at the Monaco Grand Prix as the prize.

Decision support under operational pressure

Royal Schiphol Group · Wilbur

Wilbur brings operational data, forecasts and alerts together for the teams in Schiphol’s control centres. Those teams decide under time pressure, so information has to be usable at a glance.

I owned the content end to end, from user research and usability testing to implementation.

The principle I took from it: information only becomes useful when people understand what it means and know what to do next.

Consequences before confirmation

ABN AMRO · Onboarding and Tikkie

In financial services, legal requirements, product logic and user confidence all meet in the same sentence. I worked on onboarding and payment journeys, in Dutch and English, where people had to understand the consequences before clicking or confirming.

I moved key implications earlier in the journey and made review moments clearer, so people could decide with confidence without extra friction.

UDA · My framework

Understand. Decide. Act.

Don’t make me think about the interface.
Help me think about the decision.

A flow can work perfectly and still help no one choose. The usability test passes, the button is clear, the information is correct, and people still drop out just before the choice. Or they choose, and call support two weeks later.

Most journeys design two things explicitly: information and interaction. The decision in between is assumed. UDA makes that layer visible, so teams can diagnose it and design for it. I call the quality they’re designing for decisionability: how well an experience lets someone make an informed, confident choice.

U
Understand

“Do I understand what this is and what my options are?”

Reduce confusion.

D
Decide

“Can I make a choice, and do I dare to?”

Make uncertainty and trade-offs manageable.

A
Act

“Can I carry out my choice?”

Turn commitment into action.

Where decisions break

Understanding and acting already have mature disciplines. UDA goes deep where the gap is: the decision. It looks at six conditions.

  • ClarityDo I understand what I’m choosing between?
  • ComparabilityCan I compare the options on the same terms?
  • ConsequencesDo I know what happens if I choose, don’t choose or change my mind?
  • RelevanceCan I tell what this means for my situation?
  • ValueCan I weigh whether it’s worth the price, effort or risk?
  • Agency & trustDoes the choice feel voluntary, transparent and reversible enough?

The outcome of those six is confidence. Confidence is designed, not declared.

The line I don’t cross

A model that explains what people need to commit can also be used to push them. So UDA has a firm boundary: an intervention must reduce the real uncertainty of a decision, not just the perceived uncertainty. Scarcity, social pressure and clever framing fall outside the model, even when they convert.

Read more

I write about UDA and the decision layer on LinkedIn, with examples from real journeys.

Read my UDA articles on LinkedIn →

UDA isn’t defined by the industry. It’s defined by the weight of the decision. Three questions fit into any design review; the depth is there when the stakes are high.

UDA is a working model in validation. The full working document, with diagnostics, research methods and measurement, is available on request.

About

I’ve never been particularly good at staying in one box.

I started in advertising: copy, concepts and creative direction for brands like LeasePlan, Gall & Gall, ICI Paris XL and Schweppes. As digital changed the work, I moved into UX writing and content design, for organisations including Eneco, ABN AMRO, CarNext, Royal Schiphol Group and Flying Blue.

Now AI is changing the work again, which suits me fine. I’ve always been more interested in the problem than in the job title. What I’ve really been doing all along is working out what we’re trying to say, and whether there’s a better idea underneath it.

Today I work between creative strategy, content experience and content systems. I like zooming out to the concept and then all the way back in to a button, a subject line or a single word. That range is the work.

A little more personal

Advertising taught me to look for the idea. Journalism taught me to keep asking questions, first as editor-in-chief of Status Quo, the law faculty magazine at VU Amsterdam, and today as volunteer video editor-in-chief at StudioDMN. UX taught me to look at what people actually need, rather than what an organisation wants to say.

I still like making things myself. More and more, I also enjoy shaping the direction around them, so good work keeps happening after I leave the room.