SpaceX: Ookla test shows direct-to-cell only fills gaps

text delivery rate 86% against 98.1% on the ground, median delivery time 16.6 seconds against 1.3 seconds, and median signal-to-noise ratio of 3.0 dB, below the 5 dB starting point for coverage in the FCC's new proposal. Yesterday's reported approach to Boost Mobile depends on next-generation direct-to-cell satellites (which need Starship) and on EchoStar spectrum coming through. The user base and the value of filling gaps are confirmed, but pricing power has to wait for the next-generation constellation. Airlines have tied Starlink to revenue for the first time: in its half-year results on 2 Nov, Japan Airlines credited ZIPAIR's 12.3% rise in passenger revenue to Starlink across the whole fleet plus flexible pricing, without separating the two. Do not read the 73% fall in the low-cost segment's operating profit as Starlink bringing profit. The negotiating story has moved from passenger experience to reported revenue. In Synergy's third-quarter cloud market data, SpaceX/xAI enters the top ten for the first time, in sixth place with about 3% share, but this is a third-party estimate. The AI business has scaled up, but customer concentration (Anthropic was 34.5% of third-quarter revenue) remains the key risk. Starlink is swapping first-generation dish customers to the Standard kit for free. Together with dynamic pricing from 30 Nov, this clears out inefficient terminals to raise capacity use per satellite, at the cost of more hardware subsidy in the short term.