Systems and methods are described for reducing lab-to-lab and/or instrument-to-instrument variability of Multi-Attribute Methods (MAM) analyses via run-time signal intensity calibration. In various aspects, multiple MAM-based instruments each have detectors and different instrument conditions defined by different instrument models or sets of settings. Each MAM-based instrument receives respective samples and a reference standard as a calibrant. Each MAM-based instrument detects, via its detector, sample isoforms of its respective sample and reference standard isoforms of the reference standard. The MAM-based instruments are associated with processor(s) that determine, via respective MAM iterations, correction factors and sample abundance values corresponding to the sample isoforms. The correction factors are based on the reference standard, and the sample abundance values are based on the correction factors. A variance value of the sample abundance values may be reduced based on correction factors of each of the MAM-based instruments.