The 10x Genomics single cell RNA-seq method is 3' end focussed and can therefore be used to detect alternative polyadenylation (APA). We have developed a method for measuring APA, giving each gene a score between -1 and 1 representing a shift in 3' end site usage from the average. Measurement accuracy depends on the number of reads available, so precision weights (1/σ²) are also calculated for each measurement, similar to the voom method for gene expression data. Data in the form of a matrix of measurements with a matrix of precision weights can be used in a generic fashion for differential testing, exploratory factor analysis, or as part of a multi-omics analysis. The method is demonstrated on Peripheral Blood Mononuclear Cell data. APA is related cis-cell-trans-gene to expression levels, and top confidently APA genes are found.