Case study · Experience strategy
The deliverable was a website. The work was deciding what a stranger needs to understand first.
Structuring understanding, then building something that carried it.
A brand strategist needed a site. The site was never the interesting problem. The interesting problem was that the normal way of introducing someone tells you they are qualified long before it tells you whether you’d want to work with them — and no amount of visual design fixes an order that is wrong.
A person kept being introduced by what she does instead of what she’s for.
Not by a previous website — there wasn’t one. By the convention. Ask how a strategist is normally introduced and you get a title, a client list, and a set of credentials, with the reader left to assemble a person out of information.
Credentials answer is this person qualified. But that is not the question a visitor is actually holding. They are deciding whether they want to work with someone, and that decision does not run on qualifications.
Title, then work as proof of the title, then a paragraph about the person, then contact. Credibility first, on the assumption that credibility is what convinces.
Credibility answers a question nobody has asked yet. Someone has to care first. So the order was built around how a stranger actually comes to trust a person.
Sequence the site around a decision, not a content inventory.
Every section was assigned a single job, and the order was set by what a visitor needs to understand before the next thing can land: question → evidence → belief → process → person → contact.
Belief before biography. Learn the person first and the work reads as a résumé — a list of things that happened. Learn the question first and the person becomes the answer to it. Same sections either way; the order decides which one someone leaves with.
Five decisions, all still visible on the live site.
Each one could have gone the ordinary way. None of them are styling — they are choices about what a visitor is allowed to understand, and in what order.
The header identifies. The narrative introduces.
The descriptor exists — it just isn’t responsible for creating understanding. It sits small and functional beside her name, doing a header’s job, while the first thing said at full size is what she’s for.
On the siteTurning stories into something people want to belong to.
The work section claims a pattern instead of listing projects.
Six projects across entertainment, publishing and events could easily read as a varied client list. Named as a pattern instead, they become evidence for one recurring instinct — which is the actual thing being sold.
On the siteI’ve noticed I tend to chase the same kinds of questions, even when the projects look completely different.
The idea leads. The client is metadata.
Recognisable names are present, but underneath. Each project is introduced by what it was trying to do, so the work argues for the question rather than borrowing status from a logo.
On the siteMaking adaptations feel like something fans could be part of.
Don’t make someone leave the story to experience the story.
Her process content lived as links out to Instagram. Functionally that worked; experientially it ended the visit — one click and the visitor was inside an infinite feed, in someone else’s environment, with no reason to come back. So the content was rebuilt to feel native to where it came from while staying inside her site. The decision was about custody of attention, not about embedding.
The ruleThe story shouldn’t be somewhere else.
No résumé anywhere — and the close returns to the open.
No dates, no employers, no job history on any part of the page. Nothing to skim, nothing to filter by. And the last line answers the first: the site opens on belonging and ends on it, so it reads as one argument rather than a set of pages.
Opens…something people want to belong to.ClosesLet’s build something people want to be part of.
It stopped being a résumé and became an introduction.
Jane is my wife. This was personal; the discipline was real. She was the stakeholder and the subject, and the job was never to invent a persona — it was to make what was already there legible. The only real pushback through the build was hers, and it was the right one: keep making sure this feels like her and not like a professional template.
The reason this sits beside the enterprise work rather than in a separate design folder is that it is the same move, performed on a person instead of a process. Called experience strategy here, systems work there. The scale changes the name and almost nothing else.
AI didn’t shorten the thinking. It shortened the distance between a decision and something you could react to.
- Identified the problem
- Shaped the experience strategy
- Made the UX decisions
- Directed the visual system
- Tested each iteration
- Gave the feedback that changed it
- Stakeholder
- Subject matter expert
- Owner of her brand voice
- Final approval on whether it sounded like her
- Closed the gap between deciding and seeing
- Turned decisions into working experiences
- Made ideas testable quickly enough to be worth testing
Find the friction. Design the thing that removes it. Use whatever makes it real.
Concept → prototype → test → refine used to run at the speed of a build cycle, which meant most experience decisions were argued about rather than tried. The distance between deciding something and seeing it is now short enough that trying is cheaper than arguing. What the loop asks for is someone who knows which decisions matter and can tell when one of them is wrong.
Which is the same thing the enterprise work asked for, and the same thing Barry asked for. The tools changed how fast a decision becomes something you can react to. They did not change who has to make it.
Experience strategy · Narrative architecture · Brand strategy · UX · Visual system · Implementation collaboration
Every decision above is verifiable at janeklee.com · structure and copy recorded 18 Aug 2026 · no prior public site existed, so nothing here is framed as a redesign · no metrics claimed