Restructuring how a bank sells credit cards

The situation

Credit cards are one of the most contested product categories in Australian banking, and one of the most heavily analysed. Everyone had opinions about the pages. What the squad didn't have was a clear picture of what customers actually did on them.

The decision

Before proposing anything, I built the evidence base: traffic flows across the category, and HotJar heatmap analysis of how people actually moved through comparison and product pages. The pattern was consistent. Customers weren't failing to understand the products. They were failing to find the right one, backtracking through a structure organised around how the bank thought about cards rather than how a person chooses one. The intervention I proposed wasn't a rewrite. It was an information architecture restructure: reorganising the category around customer decision paths, then rewriting only what the new structure demanded.

What shipped

A restructured credit card category: new hierarchy, new comparison logic, rebuilt page structures, and plain-language content where legal complexity had been hiding the value proposition. All of it through the approval reality of a big four bank, where every change touching a regulated product clears risk, legal and compliance before it ships. That constraint is the job, not an excuse. Work that only converts in a wireframe isn't finished.

Proof

What it says about how I work

Information architecture is where I look first, because structure fails silently. Copy problems announce themselves; architecture problems disguise themselves as everything else, including as copy problems. The discipline is to instrument before you intervene. It's slower for the first fortnight and faster for every month after.

Next case study: two government websites, one statutory deadline →