Increasing cost differentials between modes of data collection and countries are requiring users and practitioners to consider more cost-effective survey designs. Using a 'fitness for purpose' framework, the argument is made that the tools exist to enable objective evaluation of alternative designs using a variety of methods within a common framework that can be shared by all survey users. The paper argues that coverage will be one of the largest sources of potential bias in any survey using data-collection methods other than face-to-face or mail. The calibration of coverage is therefore a pre-requisite in any discussion of alternative survey designs. The paper discusses the need for clear, unambiguous definition of the characteristics of mode access to reflect the realities and constraints of research practice. It illustrates this with a discussion of internet access from home, suggesting that some current commonly used definitions are misleading as to the true level of coverage. The paper comments on the general lack of reliable coverage data for new modes. Using data from Eurobarometer and from the USA National Health Interview Survey, it tabulates the coverage of the internet, mobile telephones and sub-categories of technology in Europe and the US. The paper argues that these data show differences that are so considerable that in many countries it would be ill-advised to use single-mode approaches to achieve reliable estimates. The paper concludes that, to provide cost-effective international measurement, we must embrace the use of mixed-mode methods within countries. This will require a broad international effort to agree definitions, measurement of coverage, and better questionnaire design, to reduce/eliminate mode effects.