High-resolution mass spectrometry is the foremost technique for qualitative and quantitative lipidomics analyses. Glycerophospholipids and sphingolipids, collectively termed polar lipids, are commonly investigated by hyphenated liquid chromatography-mass spectrometry (LC-MS) techniques that reduce aggregation effects and provide a greater dynamic range of detection sensitivity compared to shotgun lipidomics. However, automatic polar lipid identification is hindered by several isobaric and isomer mass overlaps, which cause software programs to often fail to correctly annotate the lipid species. In the present paper, a buffer modification workflow based on the use of labeled and unlabeled acetate ions in the chromatographic buffers was optimized by Box-Behnken design of the experiments and applied to the characterization of phosphocholine-containing lipids in human plasma samples. The contemporary generation of [M + CH3COO]-, [M + CD3COO]-, and [M - CH3]- coupled with a dedicated data processing workflow, which was specifically set up on Compound Discoverer software, allowed us to correctly determine adduct composition, molecular formulas, and grouping, as well as granting a lower false-positive rate and streamlining the manual validation step compared to commonly employed lipidomics platforms. The proposed workflow represents a robust yet easier alternative to the existing approaches for improving lipid annotation, as it does not require extensive sample pretreatment or prior isotopic enrichment or derivatization.