Research background and purpose:
Quick commerce - delivery within 60 minutes - has expanded rapidly across digital retail, but margins remain thin where delivery fees compete against dense networks of physical stores. Operators charging these fees need a way to identify which consumers will pay them, using only the data the platform already has. This study tests whether mobile operating system (iOS vs. Android) serves as a valid segmentation criterion in the Polish q-commerce market. No prior study has tested the OS-as-proxy hypothesis in a grocery or delivery-fee context. Several operators have already exited the Polish market, and a segmentation strategy that costs nothing to implement - OS is already visible at checkout - has direct implications for surviving operators.

Design/methodology/approach:
A Computer-Assisted Web Interviewing (CAWI) survey of 1,000 mobile users in Poland (2025) was used to test three hypotheses concerning iOS/Android differences in willingness to pay for standard delivery (H1), speed surcharges (H2), and the independence of any OS effect from household income (H3). Data were analysed using Mann-Whitney U tests and ordinal logistic regression with demographic controls.

Findings:
H1 was not supported in either group - the estimated iOS effect is directionally consistent with the hypothesis but statistically inconclusive given limited power. H2 was rejected: the only significant OS effect runs counter to the prediction, with Android users showing higher willingness to pay for faster delivery than iOS users. H3 was confirmed: OS and declared income are statistically independent, ruling out income confounding and pointing toward service context as the primary boundary condition.

Value added and limitations:
OS may be a structurally weak income proxy in markets where Android spans the full device price spectrum. Device value should be tested as a potentially more appropriate segmentation construct. The study is limited to Poland, a market with exceptionally high retail density, and employs a cross-sectional design.