Every advertiser asks a version of this, and most of the answers they get are marketing. Here is the honest one, including the parts that do not favour us.
The objection is correct
A rewarded user did not discover your app. They wanted a currency, your offer was the route to it, and they did the minimum the offer required.
Everything that follows from that is real:
- Retention is lower than from a channel where the user chose you. Often substantially.
- Early events over-convert. Installs and registrations look spectacular, because completing them was the entire point.
- Intent is thin at the top. A rewarded install tells you almost nothing about that person's interest in your product.
Anyone who tells you otherwise is selling. Price the channel as what it is.
What is also true
These are real people doing real actions. They installed the app, they registered, they reached level 5. That is different in kind from bot traffic, click farms and attribution fraud, which is what "low quality" usually means elsewhere.
The volume is immediate and controllable. A campaign can be live and delivering within a day, in specific countries, on specific devices, with a hard budget. Few channels do that.
Deep events survive the filter. This is the part most advertisers miss. A user who reached level 5 for coins still reached level 5 — they experienced your product properly. The proportion who stay is lower than from organic, but it is not zero, and you knew the price in advance.
You choose the depth. That is the whole lever. If early events over-convert, stop buying early events.
Where it genuinely works
- Anything with a funnel you can price at the bottom of. Buy the deposit, the purchase, the day-seven engagement — the users who get there are worth having, whatever brought them.
- Games with a real early experience. A player who is enjoying it at level 5 does not stop enjoying it because coins brought them.
- Filling a cohort. New territory, a seasonal push, a soft launch that needs volume — rewarded delivers scale on demand.
- Surveys and research. The panel industry has been paying for attention for decades, and it is the correct model for it.
Where it does not
- Brand campaigns. If the goal is sentiment or awareness, paying someone to look at you buys neither.
- Anything priced on an install alone, unless you already know your funnel cold and have priced the drop-off in.
- Products with a long, high-intent consideration cycle. Nobody chooses a mortgage for coins.
- Vanity metrics. If the number you report is installs, this channel will make that number beautiful and your business worse.
How to find out for yourself
Do not argue about it — measure it, and measure the right thing:
- Tag the cohort so rewarded users are separable from every other source in your own analytics.
- Compare them at day 7 and day 30 against your other channels, on revenue rather than on activity.
- Work out your effective cost per retained user, not per install.
- Then decide. If it clears your bar, buy more. If it does not, move the goalpost deeper and price it again before writing the channel off.
The summary
Rewarded traffic is real people, incentivised, at scale, on demand, with a known and controllable acquisition cost — and with materially lower intent than any channel where the user came to you first.
It is an excellent volume channel bought at the right depth, and an expensive way to inflate an install count bought at the wrong one. Which of those you get is decided by the model you pick and the price you set, not by the channel.