Yeah, I'm skeptical, too. Especially of surveys which ask respondents about the future. I.e., here they asked them if they used, or _planned_ to use whatever cl

Yeah, I'm skeptical, too. Especially of surveys which ask respondents about the future. I.e., here they asked them if they used, or _planned_ to use whatever cloud platform. This is problematic because you rarely get good results from asking people about what they might do (e.g., choose to use Azure over AWS). And like you say there is likely to be bias which relates to the composition of their survey sample. All in all, the more accurate conclusion is that for those surveyed there was more interest in Azure over AWS (since the way there questions were framed it's very hard to even tell exactly what proportion are already using one or the other). Interest may, or may not, map to future choice. Respondents may signal interest in Azure, yet pick AWS when the rubber hits the road.