Recently major transport operators such as Transport for London or RATP in Paris developed an interest for walking issues, and seek to promote walking as an integrated part of their transport offer. This paper draws from two case studies being carried out on behalf of these operators. It will explain in which context this interest has risen, what are the motivations behind the development of a walking-friendly network (focusing on its hubs, their attractiveness and accessibility as well as their relations to other attractors), and how targeting the walker as a potential client for public transport can be achieved. The paper will show how walking and public transport enhance each other as they enrich the realm of possibilities regarding both mode choice strategies and route strategies. It will also explore walkings pivotal role in associating public transport with individual modes of transport. Walking performance is decrypted as a vital factor in developing full-fledged multimodal trip-chaining capacities, especially when linked with on-demand modes of immediate and free access, such as vélib and autolib. A typology of walking facilitators and hindrances will be brought forth to describe both optimal and suboptimal conditions within the exchange nodes and interfaces themselves as well as beyond into the public realm. This paper will also describe the advantages and the shortcomings of a certain number of design principles aimed at promoting walking, for instances active frontage that facilitates navigation, pedestrian interfaces that fulfil pedestrian-specific needs, and the weaving of over & underground spaces into a continuous walkscape. Finally, this paper shall deal with the crucial notion of which are the optimal scales to develop an operating walking network in order for walking to be truly efficient, not just by itself but especially in combination with other modes. The paper outlines the idea that the multimodal walker is at its best at the agglomeration scale, and goes on to give some groundbreaking arguments on why this is so.