Modern fisheries assessments increasingly rely on high-precision abundance data and indices produced by fisheries-independent surveys. Further, advancements in underwater optical technology has allowed for the simultaneous collection of both fish and habitat data. In theory habitat data can be used as covariates to explain fish abundance trends but it is often the case that individual metrics (e.g. areal cover of coral) do not explain trends and often only marginally improve precision. In contrast, aggregative habitat-complexity metrics have shown improved explanatory capacity in this regard. Herein we compare two approaches to constructing habitat-complexity metrics based on their ease of use and relationships to fish abundance and diversity. The visual habitat-complexity metric, derived from a visual scaling procedure, proved to have the best capacity to explain both fish abundance and diversity. Conversely the habitat diversity metric, estimated using Shannon-Weiner equations, allows for quick creation of a metric from historic data, showed less powerful but similar relationships in comparison to the habitat- complexity metric. We recommend that video based surveys include some form of habitat complexity data during video annotation as the approach was efficient in explaining fish abundance and diversity trends. Specific use of either method demonstrated here will depend on the state of historic data, staffing, capacity to annotate video, and time constraints.