The fundamental distinction between samples is whether they are probability or non-probability samples. A probability sample is often a random sample and only this kind of sample can be relied upon to be without bias. The simplest form of non-probability sample is a convenience sample. The second form of non-probability sample is a judgement or judgemental sample. The third type of non-probability sample is a quota sample, and this is often used to survey large populations. The sample must be randomly selected if it is to be without bias. Since surveys are not completed by companies but by individuals, if supplier operate in a business-to-business market they have to decide which individuals. Very often the individuals are chosen on the basis of convenience - the people most commonly dealt with whose names are readily to hand. The chapter explores the problem of 'non-response bias' can be a considerable one in customer satisfaction measurement studies.