Abstract Purpose Gait functions decline with aging, characterized by shorter strides, reduced hip range of motion and increased cadence and variability. Wearable biofeedback devices have been shown to improve gait, such as lengthening strides in older adults. While these devices attempted to improve the entire gait by suggesting changes in one gait parameter, it was not known if this approach could change gait patterns differently among elderly people. This study explored how a smart insole providing biofeedback on swing time only could produce different gait patterns. Methods Thirteen older adults walked on a flat walkway with and without biofeedback from a smart insole to increase the swing time. Gait analyses were recorded immediately after each condition using the GaitRite and Xsens systems. The sample was then divided into Group A (n = 7) who increased their stride length and Group B (n = 6) who did not to analyze within-group gait changes. Results Baseline analysis confirmed no significant gait differences between the two groups. Upon using the biofeedback devices, all participants increased over 5% their swing time and Group A significantly increased their hip extension (18.01%), hip flexion (5.1%) and stride length (5.5%). Conversely, the device caused Group B to significantly reduce their stride length (3.6%), velocity (13%), and increased gait variability (stance time COV 112%). Conclusions Biofeedback suggesting instant modifications in one gait parameter could produce two distinct walking patterns in participants with similar walking characteristics. Caution in biofeedback device design and gait analysis should be taken as gait outcomes could vary among users.