nathanhaslewood.com.aumelbourne, australia
Every digital experience now has three readers.
People deciding whether to trust you. Google deciding whether to rank you. AI assistants deciding whether to recommend you. I build digital experiences that win with all three, in industries where getting it wrong is not an option.
Reader one: a person. That's you.
This page was written for you first. Everything on it is a verifiable claim about work I have shipped: product pages that still rank first in Google two years on, information architecture that moved bank revenue, production AI systems running without me.
Now try the other two tabs. They rebuild this page, live, the way a search crawler and an AI assistant each receive it. Most corporate sites fail those two views. This one is built to pass them.
This site also publishes /llms.txt and a page written for you, if you're the third reader.
Every panel above is generated at runtime from this page's actual markup. Nothing is mocked. How this site is built for three readers →
Estimated revenue contribution from two product pages I built at ANZ
Google position those pages still hold, two years after shipping
Articles a month produced by AI systems I designed and handed over
Major digital deliveries shipped on time and to budget across finance, government and defence
/work/led and delivered
Decisions that moved numbers
All eight case studies →The brief said fix two pages. I rebuilt the category.
ANZ's commercial lending products were invisible in search. I made the case for two new product pages, shipped them, and set the template the bank still uses.
Restructuring how a bank sells credit cards
Heatmaps, analytics and A/B testing showed customers fighting the information architecture. The restructure lifted conversion 10 to 15%.
Two government websites, one statutory deadline
Merged two complex Victorian Government sites into land.vic.gov.au, through a sudden shift to remote delivery, without missing the date.
/systems/designed and built
Production AI, with governance built in
Inside the systems →A content engine that runs without me
Built for a Melbourne agency: topic discovery and writing for 50+ clients across seven verticals, voice-matched per client, with two passes of critical self-review and factual verification before anything ships. 150+ articles a month, designed for zero-handover operation by non-technical staff.
Rankline, my own product
A lead engine for Australian truck and equipment finance brokers: hub-and-spoke page systems, verified finance facts, content built for all three readers. A niche deliberately distant from my employed work, where I test the method with my own money on the line.
/about/the evidence wall
Receipts, not adjectives
The full story →
/writing/the thinking
Ideas I'll stand behind
All writing →Your digital experience has three readers now
People, Google, and AI assistants judge your experience simultaneously. Most organisations are built for the first, tolerated by the second, and invisible to the third.
Number one in Google, invisible to AI
A brand can dominate search and still not exist to AI assistants. The mechanism is client-side rendering, and the test takes two minutes.
/book/shipped, cover to cover
Build, Stabilise, Leverage
A 441-page guide to building wealth through Australian property, written from my own two builds, one renovation, one sale, and every expensive mistake along the way. Researched, written, designed, published and marketed solo. And the second book, How to Buy Property With Your Super, is already free in full.
One of the best team leads I've worked with - always bringing an organised approach to leading his team.
/contact/
Work with me
Alongside my full-time role I take a small number of advisory engagements: AI-era discoverability diagnosis, content system architecture, and digital delivery in regulated environments. If your brand ranks well but AI assistants never mention it, that is usually fixable, and usually cheaply.