For the last few weeks I've been running PropellerAds, RichAds, Adsterra, and HilltopAds all at the same time. Clicks are stable, sure, but engagement? It's patchy. not where it should be given the traffic volume, and the real killer is variable isolation - can't tell if the problem is the traffic itself, the offer match, the landing pages, or just a lack of statistical significance when everything's moving at once.
Starting to wonder if running everything in parallel actually speeds things up, or if I'm just burning budget on expensive confusion.
for those who've gone this route: did you lock onto one network first, or did you genuinely get faster results with the shotgun approach? How many rounds did it take before you had something repeatable?
oh, and I'm already poking around for more push networks - any other names I should add to the list?
here's what I've been hearing from the community:
📌 Pick one network, one or two offers, one lander - push real volume until you've got clear stats, then tweak. Testing everything at once makes it impossible to know what works. Also look into recurring software affiliate programmes, one solid converting offer can become a very good living.
📌 Parallel testing speeds up learning, but only if you lock down variables hard. If you're mixing multiple push networks, offers, and landers at the same time, you nearly always learn nothing because every failure has four possible causes. Cleaner approach: 1 offer, 1 lander, 1 angle - then test source A vs source B. Once one source shows stable CTR and downstream engagement, start rotating landers or offers. Otherwise you're just deciding whether the problem was traffic quality, audience intent, creative mismatch, or weak post-click conversion.
parallel testing is useful - but only when the rest of the stack is frozen. If everything's moving, it's usually just expensive confusion.
