After setting each up with an account, paying for service and exploring how each works, I connected them concurrently to a Microsoft Surface 3 tablet, a second-generation iPad Mini, a Google Nexus 7 tablet and a third-generation MacBook Air. Then I gauged how long each lasted on its battery. After connecting with a client, I started a stopwatch and ran YouTube videos continuously until the hotspot's battery was empty. I repeated this three times, averaged the results and rounded to the nearest five-minute interval.
Next, I measured their Wi-Fi range by connecting with a client and starting up an Internet radio station. As I walked away from the hotspot, I listened for break-ups and stutters while monitoring the system's wireless signal strength. When it disconnected, I noted the place and moved back towards the hotspot. After it reconnected, I confirmed the place where the client lost contact.
I tested each hotspot's performance at several locations side by side using Ookla's Speedtest.net website. Each data set was completed within 15 minutes to lessen any effect of Internet congestion. I took daily readings at an office in Westchester County (just north of the Bronx) twice a day, three times a week. Then I hit the road and made day trips to New York City, New Jersey and Connecticut, taking readings three times a day for two days a week. Finally, I went on a four-day business trip through Massachusetts, New Hampshire and Maine. The hope was that testing the hotspots at a large number of different locations at different times evened out any high or low spots.
In addition to recording the latency (through the ping test), I tracked each hotspot's download and upload speeds. All told, I took more than 1,500 readings at 20 different locations over six weeks of testing.
Sign up for CIO Asia eNewsletters.