<p>Order picking, the process of retrieving products from their storage location, is typically performed by humans and is recognised as the costliest intra-warehouse process. A substantial portion of order picking time is spent travelling between product locations, making the prevailing routing policy a key driver of overall warehouse productivity. Warehouses may apply either an optimal routing policy, which ensures the shortest path, or more intuitive routing heuristics. While most studies have approached this topic from a deterministic perspective, few studies recognise an important human factor: workers do not always strictly adhere to routing instructions and may deviate from them. Existing research suggests that optimal routing is robust to these deviations and is therefore preferred. We revisit this claim using an extensive agent-based simulation model that better reflects contemporary warehouse practices and interactions between picking policies, for example, by applying routing heuristics in their most effective storage environments. Unlike prior studies, our results suggest that routing heuristics are viable alternatives to optimal routing in terms of performance. This further strengthens their potential for adoption by practitioners, as earlier research has already highlighted that they impose lower cognitive and mental demands on human operators. Our findings nuance the belief that optimal routing should always be preferred, and prove that human-centric modelling approaches are essential to recognise human behaviour and still achieve high operational performance.</p>