<p>Additive manufacturing of metal components enables streamlined production but less consistent material properties relative to traditional batch manufacture, due to spatial variations in microstructure. Measuring these inconsistencies is important, both from part to part and in terms of spatially variable properties within a single component. Plasticity within localized neighborhoods of grains (around 0.1–1&#xa0;mm in length scale) can influence part performance, especially when coincident with geometric stress risers. Suitable testing methods must strike a balance between throughput and precision to produce statistically significant sets of measurements. Fortunately, arrays of tests present an opportunity to leverage residual analysis, a group of statistical methods employed to discover trends and factors not already captured by a model. Here, residual analysis and uncertainty quantification (UQ) are connected to estimate profilometry variations originating from distinct sources of uncertainty. Around 100 profilometry-based indentation plastometry measurements were gathered from a cold-sprayed additively manufactured Al 7075 specimen. Four variability sources were considered, small tilts in the sample surface, small variations in indentation load, unspecified imprecision in the test method, and variations in the effective properties of the sample between different testing locations. Using residual analysis, linear regression, and principal component analysis, the unique contributions of the different uncertainties were approximated. The precision of the testing was sufficient to detect the spatial variability of cold-sprayed Al 7075 plasticity at the 1&#xa0;mm scale. Significant uncertainty was attributed to nuisance factors. Initial methods for source-specific UQ were employed to isolate material variability in the test array, with a relative precision improvement of 20% over standard practice. Connections are made to the state of mechanical property standardization in metal additive manufacturing.</p>