In conservation science or applied ecology there is often a range of techniques available to study a given problem. It is clearly sensible to use the most effective, but the cost-effectiveness of different methods is rarely compared. Comparisons are not straightforward, as the selection of the best method will depend on the time, budget and equipment, and human resources available. The most efficient method in one case might not be that for another with different constraints. To overcome this problem we use a sub-sampling approach to make such comparisons among research techniques, comparing the results achievable with increasingly larger subsets of data, and using a cost-efficiency analysis. We illustrate this approach with a comparison of two common techniques, radio-telemetry and transects surveys, in the study of habitat selection by the lesser kestrels Falco naumanni. Habitat preferences were determined using compositional analysis with the two methods providing similar overall results. Telemetry resulted in a larger number of significant differences between the habitats, but the costs were higher; 312 EUR per statistically significant difference compared to 82.5 EUR for transect data. Transects were a cheaper technique but when surveying a large area they are limited by road availability. We suggest that the comparison of the effectiveness of techniques should be routinely applied to a range of conservation issues.