Product-market fit is elusive in general, and acutely so on mobile, where distribution pipes are either constrained or flooded. I’m seeing too many teams building for Android too early. Unless there is a huge foundation under the iOS apps, building for Android is likely only to result in a few spikes in user growth and then a lifetime of hair pulling — too much for a small startup to handle.
This is going to upset Android loyalists but but I think Semil’s right here. For most SV startups and those targeting primarily a Western audience, this is all true. You can always start with a baseline of iOS users and then scale from there, even if it’s a social/communications app which, at scale, you anticipate having say about 75% of your users running Android. The only exceptions to this might be localized apps that are targeting a developing market in particular, in which case the amount of iOS users will be so small as to be meaningless for testing purposes.