The paper will discuss the challenges the national statistical organizations are facing when using the combination of web and interview data collection. Two continuous EU inquiries are compared: the EU Consumer Barometer, which is a household survey, and the Job Vacancy Survey (JVS), an establishment survey with a rotating-panel sample. For example, cost-efficiency and non-response are studied. The JVS is a mixed mode survey since 2006, whereas for the Consumer Survey the first mixed mode pilot was conducted in March 2011. For the JVS the share of the web responses at present is about one third and the overall response rate is considerably high, while the results of the Consumer Survey pilot showed a low response rate for the web-phase and a higher non-response rate for the whole data collection. Results suggest that introducing a web phase to a rotating panel establishment survey may reduce the costs of the data collection in the long run. For random sample household surveys the effect may be the opposite, making it for the respondents easier not to participate.