PeerPlay
Buying Google Play testers is reasonably safe if you are paying for real people on real Android phones who actually use your app during the closed test. It is risky if you are paying for a list of accounts that opt in and then do nothing, because your production application can still be rejected after you have spent the 14 days.
I run PeerPlay, which offers both free tester exchange and paid testing, so I have a stake in this question. I'll stick to what you can check yourself before paying anyone, including us.
Paying isn't the problem, idle testers are
Google's closed testing requirement asks for at least 12 testers opted in for 14 days in a row before a new personal account can apply for production. As far as Google's public help pages go, it doesn't say where those testers must come from.
What it does make clear is the purpose: real testing. When you apply for production, Play Console asks how you ran the test, what feedback you got and what you changed. Developers regularly report rejections saying testers weren't engaged enough, and Google doesn't publish exactly how it judges that. Twelve accounts that installed once and never opened the app again are the weakest possible answer.
Red flags in a tester service
It asks for your Play Console login. No tester service needs access to your developer account. You add testers by email or Google Group and share an opt-in link. Never hand over the account itself.
It guarantees production access. Nobody outside Google can promise that, and a service that does is telling you something about how carefully it treats the truth.
It can't say what devices testers use. If the answer is vague, assume emulators or a few phones with many accounts.
It has no plan for drop-outs. Testers leave. A service that sells exactly 12 with no replacements leaves you one drop-out away from a broken streak.
What real activity looks like
A tester who counts for something opens your app on most days of the test, on a phone they actually use, and spends at least a little time in it. That is a low bar, but many cheap offers don't meet it, because the cheapest way to fill 12 slots is to opt in once and walk away.
A good service can tell you how it checks this. On PeerPlay, for example, a test day only counts when the tester starts the test from PeerPlay, which opens your app, and stays in it for at least 30 seconds. Each counted day records the device model and Android version, and testers who join and then abandon a campaign get a visible quit counter on their profile.
None of that proves what a tester did inside your app, and no tester service can prove that. But it is the difference between a service that measures activity and one that only counts names.
Questions to ask before you pay
How do you check that a tester is a real person on a physical phone? How do you know they open my app during the test, not just on day one? What happens if someone leaves on day 9? Will testers send feedback I can use in the production application?
Also ask for a real example with numbers. Our first paid campaign, for the women's health app Nada, used a 15-tester package. 18 testers joined over the campaign, 11 completed all 14 days and 3 quit. At least 12 were active every day from day two onward, and Google then granted production access. That is roughly what a healthy paid test looks like: a buffer above 12, some drop-outs, and the count never falling below the line.
What to do on your side
Watch the opted-in count on your Play Console Dashboard every day, whoever brings the testers. If it drops toward 12, get replacements before it falls below.
Check your own analytics. If 15 people are supposedly testing and your app shows two sessions a day, you are paying for names, not testing.
Read and answer feedback, fix what you can, and write it down. Those notes are the backbone of your production application.
Free options exist too
Paying isn't the only way. On PeerPlay's free route, Android developers test each other's apps one for one, opening them from PeerPlay every day on real phones, and emulators aren't allowed. Paid packages are there for developers who don't have time to test other apps.
Either way, judge any service, free or paid, by the same test: real phones, real daily use, and a buffer above 12.
FAQ
Is it against Google Play rules to pay for testers?
Google's public help pages don't say where testers must come from. What matters is that they are real people who actually test your app during the 14 days.
Can bought testers get my production application rejected?
Inactive ones can. Developers report rejections for testers who weren't engaged enough, so make sure paid testers really open and use the app.
Should a tester service ask for my Play Console login?
No. Testers are added by email list or Google Group and join with your opt-in link. Never share your developer account.
How many paid testers should I get?
More than 12. Some testers drop out during any test, so a buffer of a few extra keeps you above the minimum every day.
How can I tell if paid testers are real?
Ask what devices they use and how daily activity is tracked, then compare the tester count with your own analytics during the test.