[marianminds]

I've used numbeo before and I had a look at this, comparing about 10 or so different cities. The format and function is pretty much the same and I use them both for the same purpose -- a simple comparison of aggregate data on different cities using the same metric. I moved to Rome a year ago and in another year I'll be moving to some German-speaking country, so I'm comparing different cities now. It's not important to me whether the rental estimates are 100% correct or slightly high/low, what's important is that the data is collected the same way for both cities under comparison. I want a broader view to narrow the search, and then I will go elsewhere to get more accurate info -- apartment searches in the respective cities, for example. If you get enough data from varied enough sources to minimise biases (e.g. expats might live in the more expensive parts of the city and not know the local tricks, or most of your contributors might only be from a certain subset of expats) then I don't need to know the comparison of a hundred different types of little things because I can get more exact information elsewhere. What I'd recommend doing instead of narrowing your data selection is to actually broaden the scope of your comparison -- compare aggregated measures of quality of life, average temperatures/rainfall, hours of sunlight per day, number of bars or gyms per capita, etc. If I know what city I'm looking at, I can easily find a list of apartments for rent and get direct information that way. What I can't do as easily is compare the general perspective of life satisfaction or public transport penetration, crime stats, etc.