The GA4 segment showed one desktop visitor. Thirteen sessions, sometimes more. They landed on the homepage, navigated to /work/, and read it for seven straight minutes. They worked through two courses. Several articles. Then they left.
No conversion event. No form submission. No email.
The site had held that visitor’s attention for a cumulative stretch most sites would envy. Then it offered them nothing to do with it. No door to the next step. Just the back button.
That was the observation that opened the conversion week.
What the site was
captainrandom.co.uk had content. It had craft. It had a design system, a devlog, an /about/ that was honest. What it didn’t have was a funnel.
This is the portfolio trap. The portfolio frame asks one question: does the work look good? It optimises for impression. Content is curated; presentation is deliberate; everything is arranged for maximum credibility. And it stops there. The work is displayed. What the visitor does next is left to chance.
/security/, the security-by-design philosophy and working method, lived in the footer and a buried link in the /work/ intro. Nothing on the homepage said it existed. Nothing said anything new had shipped recently.
/work/ ended. That was it. Seven minutes of reading, and the page returned nothing. No next step, no prompt, no action point. The visitor who had been evaluating for seven minutes was left to decide for themselves what to do about it.
The timelines note on /contact/ read: “I’m not taking on rushed work. Two weeks minimum. Probably not the right person if you need something next week.” Written as a quality signal, an honest statement about how projects get done well. Experienced, through a prospect’s eyes mid-evaluation, as a rejection before contact had been made.
This is what a shop floor with no signposting and no till looks like. Stock on the shelves, customer in the doorway, no route to the counter. The transaction never happens.
The shopfront: hero band as front door
The first piece shipped was the What’s New band: a panel sitting directly under the hero.
The problem it addressed: launches had no front door. /security/ was discoverable only if you already knew to look. The hero section did what hero sections do (introduced the person and the work) but said nothing about what had changed, what had shipped, or why a returning visitor should look somewhere new.
The What’s New band announces what has shipped, where it lives, and why it matters. The blood accent at the top is the visual signal: not “here are my skills”, not “welcome to my portfolio”, but this is fresh, look here first. The shopfront in a well-run shop is exactly this: the display that tells the returning customer what’s new before they wander past it.
The aisle: /work/ as evaluation surface
The GA4 segment made the /work/ reading time legible. Seven minutes isn’t browsing. Seven minutes is evaluating. That is the behaviour of a visitor deciding whether to hire: reading case studies carefully, cross-referencing detail, forming a view. The page was already doing its job. The problem was what it offered at the end.
An aisle in a well-designed shop leads somewhere. /work/ was an aisle that ended at a wall.
The timelines note was the only copy change the aisle needed, and it was a flip rather than a rewrite. Old version: “I’m not taking on rushed work. Two weeks minimum. Probably not the right person if you need something next week.” The information is true. The framing is wrong. A prospect seven minutes into reading your work, already evaluating and already interested, reads that as a barrier rather than a credential. New version: the same information, framed as a working style rather than a condition of entry. How projects get done well, stated directly. The signal is the same. The experience of reading it mid-evaluation is not.
Out of scope: any wider /work/ copy changes. The seven-minute reading time already told me the aisle was working; the task was to add a till, not redesign the aisle.
The till: hire strip
The hire strip closes /work/. Ink-inverted panel, blood left rule, volt-coloured availability pulse. One purpose: a visitor who has evaluated and decided should have somewhere to act.
This is the till. Not a banner. Not a soft CTA hedged with “if you’d like to chat.” A deliberate close. The point in the customer journey where intent meets action and the shop makes it easy to proceed.
The volt pulse is the only animated element on the page. It signals availability. It signals now. The ink inversion separates the strip from the reading surface above it; you are no longer in the aisle, you have reached the counter. The design is doing the work a verbal “click here to hire me” never does: it marks a threshold.
Stock on the shelf: /security/ goes public
The security-by-design philosophy and working method existed internally before this week. It was a real body of work: the philosophy behind how the software is built, the architecture decisions, the working method. None of it was visible to a visitor.
The full scope shipped as a public page during the same week. Stock that had been sitting in the back room, now on the shelf via the What’s New band. The GA4 segment showed there was a visitor worth showing it to: someone who had already demonstrated they were evaluating the work seriously.
The /security/ page is not a differentiator in the portfolio sense. It doesn’t demonstrate craft to be admired. It answers a question a serious prospect asks: how does this person think about risk, quality, and working method? The answer was always there. Making it visible is a shelf-stocking decision, not a content creation one.
The retail frame as diagnostic tool
The retail metaphor earns its place because it is generative, not just descriptive.
A portfolio asks one question: does the work look good? A shop floor asks a different one. Where does the visitor go next? Different questions, different implications for every design decision. The hire strip exists because the shop floor question makes it obviously necessary; you cannot have an aisle without a till. The timelines flip exists because the shop floor question identifies it as a barrier mid-aisle. The What’s New band exists because the shop floor question asks what happens when a visitor arrives at the front door.
The same frame applies to any future addition. A new page: which part of the journey does it serve? A new feature has to move the visitor, not just entertain. Any new section belongs in one of three places: shopfront, aisle, or till. Anything that fits none of those is a display cabinet. Nice to have; doesn’t convert.
The conversion week didn’t add features. It added logic. Hero band as shopfront: the entry point that announces something is new. /work/ as the aisle: the evaluation surface the visitor had already found, read for seven minutes, and found credible. Hire strip as the till: the action point at the close.
One visitor. Thirteen sessions. Seven minutes of reading per /work/ visit. The site kept them. It now has somewhere to take them.
