The increasing availability of open access data on species occurrences is leading researchers to generate more hypotheses about patterns of species distributions. However, when all of this information is mapped onto a particular geographical scale, gaps usually appear due to lack of knowledge and sampling spatial bias (the so-called Wallacean shortfall). To overcome these problems as efficiently as possible, field surveys should be designed after distinguishing well-surveyed places from those with incomplete inventories in order to carry out the extra survey effort in those areas not represented environmentally and spatially by the well-surveyed places. This procedure requires (1) gathering, cleaning and standardizing data; (2) selecting environmental variables that are important for the group considered according to field experience and the literature; and (3) making statistical decisions about the number and location of areas that should be surveyed according to the available resources. Here, we summarize most concepts and procedures devoted to the evaluation of biodiversity data, offering some general recommendations on how to use them for optimizing new survey designs. As a practical guide for potential users, we provide an example describing its application to a comprehensive database on bryophyte distribution on Terceira Island (Azores, Portugal). More than 8,000 bryophyte records were gathered, but (i) less than half of the island area has been surveyed at least once and (ii) less than 1% of these have reliable inventories (placed on the few remnants of laurel forests that have been traditionally better surveyed). Nevertheless, surveying just 15 additional localities evenly distributed across the major environmental regions and habitats on Terceira Island seems to represent the existing environmental diversity. We believe that the survey protocol presented here for bryophytes of Terceira Island could be flexibly applied to other taxa or areas.