Positive matrix factorization (PMF) can be used to develop more targeted air quality mitigation strategies by identifying major sources of a pollutant in an area. This technique is dependent, however, on the ability of PMF to resolve factors that accurately represent all sources of that pollutant in an area. We investigated how the accuracy of PMF solutions might be influenced by monitoring data characteristics, such as temporal resolution, monitoring location, and species composition, to better inform the use of PMF in VOC mitigation strategies. We applied PMF to five VOC monitoring programs collected within a four-year period in Colorado and found generally consistent factors, which we identified as oil extraction, processing, and evaporation; natural gas; vehicle exhaust; and liquid gasoline/short-lived oil and gas. The main determinant influencing whether or not a dataset resolved each of these sources was whether the dataset had a comprehensive list of VOC species covering key species of each source. Pollution spikes were not well-modeled in any of the solutions. Hyperlocal and volatile chemical product factors expected to be resolved in the industrialized, urban location were also missing, highlighting three limitations of PMF analysis. Wind direction dependence and diurnal trends aided in source identification, suggesting that high-time resolution data is important for developing actionable PMF results. Based on these findings, we recommend that air monitoring for PMF-informed VOC mitigation efforts include high temporal resolution and a comprehensive array of VOC species.