This work examines the processing of kinematic smartphone GNSS data using Precise Point Positioning (PPP) within the framework of the Google Smartphone Decimeter Challenge 2023. Various smartphone models and different processing methods, such as a code only solution, a Kalman-filtered code solution, a Doppler- and phase-smoothed code solution, and a combined code and phase solution, are compared. The Uncombined Model serves as the PPP observation model, as it offers several advantages in the context of smartphone observations. The goal is to examine the chosen approaches in terms of accuracy and to identify limitations. Both entire measurement drives and selected route sections with varying environmental conditions are considered. The analysis focuses on a freeway drive on a straight section, a freeway intersection with underpasses, and stopping at a traffic light. The results are analyzed both graphically and statistically by comparing the determined solution trajectories with the reference trajectories of a geodetic receiver. The results show that the accuracy of position determination heavily depends on the chosen processing method and the local conditions along the route. The code solution can be significantly improved in nearly all scenarios and smartphones by applying Kalman filtering. Under simple environmental conditions, the accuracy of the solution can be further enhanced by incorporating carrier phase observations. However, in more challenging local conditions, the carrier phase tracking suffers, resulting in increased deviations in the results. The most reliable results under complex conditions are achieved with the Doppler-smoothed code solution. Under simple conditions, the phase-smoothed code solution and the combined code and phase solution are more accurate. In combination with a well tuned Kalman filter, deviations of less than one meter can be achieved.