The first step of a sampling strategy should be to clearly define its purpose. The aim typically is to obtain information for decision making. The decision rules typically involve estimates of the characteristics of the population (often the mean and standard deviation). The decision rules also require a definition of the population to be sampled. There are many techniques available for obtaining estimates of key characteristics of the population. The definition of population is critical but often nontrivial. There are many different sampling strategies available. The theoretically simplest scheme is simple random sampling and that can be used to provide estimates of the mean and standard deviation. More precise estimates can be obtained using stratified sampling. Cluster sampling is useful when there is significant travel time between the sampling units—however, except in simple cases, estimation of the mean and standard error requires expert input. Systematic sampling is simple to apply and gives precise estimates of the mean. Good estimates of the standard error are not available, but there are useful approximations.