"iOS and Android, obviously" is where most app conversations start. It is sometimes right and it is worth thirty minutes of checking rather than assuming.
First, look at your own numbers
Open your website analytics and look at mobile visitors by operating system. That number is the closest thing you have to evidence, and it varies enormously by market and by audience — the same product can be eighty percent Android in one country and the reverse in another.
If the split is close to even, both platforms are usually right. If it is heavily one-sided, launching on the majority platform first and adding the other after real feedback is a perfectly respectable plan, and it buys you a month.
What the second platform actually costs
On cross-platform, far less than people expect: one codebase, most of the work shared. The extra is store setup, platform-specific behaviours, and testing on the other kind of device — real, but a fraction of the first platform.
On native, the second platform is close to a second project. That is the case where "both on day one" deserves a hard look.
The costs nobody mentions
Each store is an ongoing relationship, not a one-off submission. Two developer accounts and their fees. Two review processes, each with its own opinions about your permissions and your subscriptions. Two sets of platform updates that will eventually force a release you did not plan. Doubling that before you know anyone wants the app is a real cost with no return.
The case for both anyway
If your product only works when both sides of a marketplace can use it, or your customers are companies whose staff carry whatever they were given, half a launch is not a launch. Then both, from the start, and budget accordingly.
The order that usually works
Ship one platform to real users. Learn what the app is actually for — it is never exactly what the plan said. Then ship the second with those lessons baked in, rather than shipping the same misunderstanding twice and fixing it twice. Both stores in the first eight weeks is still what we quote by default — the argument above is for when you want to spend that budget on learning instead.
Earlier in this series: native or cross-platform, in plain terms.