There is growing policy and regulatory interest in better aligning electricity tariffs with the cost of providing network services to customers: to provide a better price signal for economically efficient use of the network, and reduce cross subsidies between different customers. Given that network costs are significantly driven by peak capacity requirements, many proposals for more cost-reflective tariffs include a demand (capacity) component. However, there are many complexities in the implementation of such tariffs. This paper first presents a method to visually assess how cost-reflective a particular demand charge network tariffs is. We apply it to a typical demand charge network tariff proposal within the Australian National Electricity Market and actual consumption data of 3876 Sydney households, and find it to have low cost-reflectivity in terms of aligning customer bills with their contribution towards network peak demand. Such misalignment has potentially significant adverse impacts on the economic efficiency of such tariffs an issue that does not appear to have received sufficient policy attention. We then use this assessment method to demonstrate how a demand charge tariff structure can be adjusted to make it significantly more cost-reflective. This method can be applied to any tariff that includes a capacity-based component.