Most websites are organised around the company that built them.
The navigation mirrors the org chart. The homepage lists services in the order someone decided to write them down. The "About" page exists because every website has an About page. Nobody stopped to ask: who is actually going to arrive here, and what do they need in the first ten seconds?
Here's a real example of what that costs, and what it looks like when you fix it.
The brief
Pirtek is a hydraulic hose repair and replacement service operating across South Africa. Mobile workshops. 24/7 availability. Emergency on-site response. Preventive maintenance programmes. A serious, established business.
Their website was a product catalogue. Organised by service type. Navigated like a brochure. Built for someone who already knows what they need and just wants the specs.
The problem: most of their customers don't arrive in that state.
Two visitors. Same homepage. Different needs.
When you look at who actually visits a site like Pirtek's, two completely different people emerge, and they have almost nothing in common.
The panicked plant manager. Equipment failure at 3am. Production stopped. Losing R50,000 an hour. Boss calling. Not reading, scanning. Looking for a phone number, a "we're available now" signal, and some proof that help is actually coming. This person needs one thing: contact, fast.
The proactive operations manager. Calm, thorough, evaluating. Planning maintenance schedules. Building vendor relationships. Looking for preventive programmes, industry expertise, case studies, and a scheduling form. This person has time. They want to be convinced, not just contacted.
Both were arriving at the same homepage. Same headline. Same navigation. Same call to action.
One needed a phone number in two seconds. The other needed confidence over ten minutes.
The site served neither.
The principle: design for mental state, not org chart
The fix wasn't a redesign. It was a restructure.
Instead of organising the site around Pirtek's service list, we organised it around the visitor's mental state at the moment they arrive.
Crisis Mode gets speed. Emergency number above the fold. "We're available 24/7" proof. Response time estimates by region. One-click contact. No friction between arrival and help.
Planning Mode gets confidence. Preventive programme details. Industry expertise signals. Case studies with real outcomes. A scheduling form. Content that earns trust over time, not in two seconds.
Navigation follows the same logic. Primary navigation serves the two modes, "When Crisis Hits" and "Before Crisis Hits", not a list of service categories. A third persona exists too: procurement teams and technical specifiers who already know what they need. They get secondary navigation, tucked in the footer. They'll find what they need regardless. The site doesn't need to optimise for them at the expense of the two who matter most.
Why this matters before a single line of code is written
The Pirtek architecture was delivered as an interactive visual map, full persona breakdown, headline options for every major section, navigation structure, CTA logic, before any production code was touched.
That's not unusual for a project of this scale. What's unusual is doing it at all.
Most web projects start with a design. Someone opens Figma, creates a homepage layout, picks some fonts. The copy gets written around the design. The structure gets inherited from the previous site. Nobody asks whether the visitor arriving at 3am and the visitor arriving on a Tuesday afternoon planning their Q3 maintenance schedule should be experiencing the same page.
When you answer that question first, who arrives, in what state, needing what, the design becomes obvious. The copy writes itself. The navigation makes sense. And the build is faster because everyone involved knows what they're building and why.
The output
The Pirtek architecture covered:
Persona mapping: three distinct visitor types, their mental states, what they need, and which pages serve them.
Headline strategy: five options per major section, with a recommended choice starred, written to match the mental state of the visitor arriving at that page.
Navigation structure: primary nav organised by visitor journey, secondary nav for research mode, utility nav for always-available emergency access.
CTA logic: different calls to action mapped to different visitor states. Crisis Mode gets a phone number. Planning Mode gets a scheduling form. Research Mode gets a product catalogue link.
Page-by-page content briefs: what goes on each page, in what order, and why.
All of this before the first component was written.
Working on a website that isn't doing its job?
The 5-Second Clarity Checklist runs you through the same diagnostic questions I use before starting any project.
TL;DR
Most websites are organised around the company's org chart, not the visitor's mental state. This means the wrong people find the wrong information at the wrong time. The fix is to design around visitor needs (crisis mode vs. planning mode) not service categories.
Key Takeaways
- Start with who arrives, not what you do. Most sites are built for the business, not the visitor.
- Identify visitor mental states. Crisis mode (emergency) and planning mode (evaluation) need different navigation and CTAs.
- One clear CTA per page, mapped to visitor intent. Not a menu of options.
- Cut jargon and internal structure. Speak directly to the problem the visitor is experiencing.
- Measure what matters: time on page, scroll depth, conversion rates. Not vanity metrics.