Most businesses don't have a clear picture of where AI or better software would actually make a difference, so it tends to get applied everywhere at once as a general upgrade, or nowhere at all because nobody's sure where to start. I work through the business function by function, creative and design work, sourcing and production, customer-facing operations, and pin down where a genuine bottleneck exists rather than assuming one industry-wide fix will cover everything.

On the creative side, that might mean using AI to generate and test directions faster, cutting down the number of rounds it takes to land on something worth developing further, or speeding up technical work like pattern development and tech packs that would otherwise eat days of a small team's time. Further down the chain, it might mean giving sourcing or production visibility they don't currently have, replacing a process that runs on memory and back-and-forth messages with something that actually holds information and surfaces problems early.

On the customer-facing side, it can mean using data and automation to understand behaviour and respond to it faster, rather than reacting after the fact. None of this means adopting every new tool that comes out. The point is finding the two or three places in a specific business where AI or software removes real friction, and being just as clear about where it wouldn't help and shouldn't be forced in.