There are several IoT platforms providing a variety of services for different applications. Finding the optimal fit between application and platform is challenging since it is hard to evaluate the effects of minor platform changes. Several websites offer reviews based on user ratings to guide potential users in their selection. Unfortunately, review data are subjective and sometimes conflicting – indicating that they are not objective enough for a fair judgment. Scientific papers are known to be the reliable sources of authentic information based on evidence-based research. However, literature revealed that though a lot of work has been done on theoretical comparative analysis of IoT platforms based on their features, functions, architectures, security, communication protocols, analytics, scalability, etc., empirical studies based on measurable metrics such as response time, throughput, and technical efficiency, that objectively characterize user experience seem to be lacking. In an attempt to fill this gap, this study used web analytic tools to gather data on the performance of some selected IoT cloud platforms. Descriptive and inferential statistical models were used to analyze the gathered data to provide a technical ground for the performance evaluation of the selected IoT platforms. Results showed that the platforms performed differently in the key performance metrics (KPM) used. No platform emerged best in all the KPMs. Users' choice will therefore be based on metrics that are most relevant to their applications. It is believed that this work will provide companies and other users with quantitative evidence to corroborate social media data and thereby give a better insight into the performance of IoT platforms. It will also help vendors to improve on their quality of service (QoS).