We report findings of an experiment that investigates the effects of augmenting IVRs with visuals while varying menu depths and menu positions on the usage by non-tech savvy participants in rural India. For the experiment, we designed four different versions of IVRs capable of informing market rates of agricultural commodities to members of farming community, namely audio-visual-shallow, audio-visual-deep, audio-only-shallow and audio-only-deep. We asked each participant to find rates of four different commodities, out of which two commodities appeared "early" in the menu, while the other two appeared "late" in the menu. The results indicate that the use of visuals in IVRs significantly increases the likelihood of successful completion of the users' tasks even with variations in menu depth and menu position. The experiment contributes by illustrating the possibility of realizing deep menu depths in visually augmented IVRs. It is hoped that such integration would help realizing non-conventional applications of IVRs in developing regions.