West Coast Wilderness Railway: proving a redesign before a line of it is built

The Partner

West Coast Wilderness Railway runs heritage steam journeys through Tasmania's rugged West Coast wilderness.

The Challenge

For West Coast Wilderness Railway, the Rinadeena, Rack & Rainforest page carries the most booking intent. This made it the page most worth improving and the most expensive to get wrong.

• Every change to the site goes through a developer, so even a small experiment carries a cost and delay.
• A live redesign has no control group. Replace the old page and its performance over the same period cannot be recovered.
• The page with the most to gain is also the page with the most to lose.

The Data Story System

Anyone with conversion experience can argue for navigation that stays on screen, reviews near the decision, or a booking button within thumb reach. But we wanted to test this with West Coast Wilderness Railway's own visitors. We wanted to test the changes before committing to them.

So we built a complete working copy of the Rinadeena page on its own subdomain in days using Claude integrations, without a developer or a staging environment. Our current AI tooling let us redesign a live page far faster than a traditional build would allow. This made the test easy and inexpensive to roll out.

We made three structural changes to the test page: navigation that stays on screen while scrolling on desktop and mobile, a reviews section below the map, and a Buy Tickets button pinned to the bottom of the screen on mobile.

The Plan

• A full working duplicate of the page on its own subdomain. No new hosting to arrange, no staging environment to negotiate, no change to the live site.
• Split the traffic. Send half of arriving visitors to each version, so both met the same audience over the same weeks.
• Read the behaviour. Measure the share of people who clicked through to the booking site, separately on mobile and desktop.

The test ran from 23 June to 26 July 2026, 34 days, with both versions live across the whole period.

The Results

Across all devices, the test version lifted clicks through to the booking site by about 49% against the live page over the same period. Mobile rose about 64%, desktop about 39%.

Indexed vs baseline (start = 100): All devices +49%; Mobile +64%; Desktop +39%

The three changes shipped together and were measured as a set, so we cannot attribute the lift to any single one. The combination outperformed the current page with West Coast Wilderness Railway's own visitors.

On that evidence, we made the recommendation to apply all three across the site. Thanks to our ability to build subdomain pages with an AI plugin, the evidence was produced without a developer and without touching the live site.

Key Takeaways

• A full working test page can be stood up in days, so a redesign is proven on real traffic before it is funded.
• Test, measure and decide before committing developer cost.

Services

Conversion rate optimisation · Experiment design and analysis · Analytics and measurement · Website performance reporting