Since we're collecting data from the actual user location, we really do capture the true experience of coverage - if a busy, popular, location has good coverage

Since we're collecting data from the actual user location, we really do capture the true experience of coverage - if a busy, popular, location has good coverage that will be reflected already in the data. This means that a large country could do extremely well, provided it covered the important urban locations. It obviously still takes more effort on the part of the operators to provide this coverage in a large country, but we're less affected by this problem than more traditional geographic coverage approaches. Although, that said, I think it could still be interesting and it's definitely feedback we'll take on board for next time. I'd say that the gap between 50% and 80% time on LTE is actually pretty important to the user.