Kike Hernandez
Newbie
- Mar 30, 2026
- 18
- 6
Running MVAS in TH/MY/ID/PL/TR. Spent the first few months losing money on bad sources. Built a vetting process that cut my wasted spend by about 70%. Sharing it here.
Why vetting matters more in MVAS than other verticals:
In most CPA verticals, a bad traffic source just means low CR. In MVAS, a bad source can:
My pre-spend checklist:
Before I run a single dollar on any new source, I ask:
1. Does the source support carrier-level targeting?
Not GEO-level. Carrier-level. If the answer is no, I don’t run MVAS on it. Full stop. The performance difference between carriers in the same country is too large to ignore.
2. What’s the mobile data vs WiFi split?
MVAS carrier billing requires mobile data. Any source that can’t tell me their mobile data percentage or won’t share it is a red flag. I want 70%+ mobile data minimum.
3. What’s their experience with subscription offers specifically?
“We run mobile offers” is not the same as “we run carrier billing subscription offers.” Ask specifically. If they’ve never run WAP billing flows before, the learning curve will cost you money.
4. Can they provide sample traffic data from similar campaigns?
CR benchmarks, approval rates, carrier breakdown. Any legitimate source running MVAS traffic will have this. If they can’t or won’t share it, move on.
5. What’s the minimum test budget and can I cap daily spend?
I never start without a hard daily cap. $50-100 max for day one. If they require a large minimum commitment upfront, that’s a red flag.
The $100 test protocol:
If a source passes the checklist, I run exactly $100 with these parameters:
Green on all three → scale to $300/day
Yellow on approval rate → pause, investigate source composition
Red on approval rate → cut immediately, don’t optimize
The ongoing monitoring cadence:
Even sources that pass the test degrade over time. My weekly check:
Running this process across 5 GEOs. Always looking to expand the source pool with quality in-app and DSP inventory. If you’re running traffic in these markets and want to compare notes on what’s holding quality, I’m genuinely interested in the conversation.
Not here to sell anything. Just think people running the same GEOs should talk more.
Why vetting matters more in MVAS than other verticals:
In most CPA verticals, a bad traffic source just means low CR. In MVAS, a bad source can:
- Tank your approval rate (network stops paying)
- Get your offer flagged for fraud
- Get your account reviewed or suspended
My pre-spend checklist:
Before I run a single dollar on any new source, I ask:
1. Does the source support carrier-level targeting?
Not GEO-level. Carrier-level. If the answer is no, I don’t run MVAS on it. Full stop. The performance difference between carriers in the same country is too large to ignore.
2. What’s the mobile data vs WiFi split?
MVAS carrier billing requires mobile data. Any source that can’t tell me their mobile data percentage or won’t share it is a red flag. I want 70%+ mobile data minimum.
3. What’s their experience with subscription offers specifically?
“We run mobile offers” is not the same as “we run carrier billing subscription offers.” Ask specifically. If they’ve never run WAP billing flows before, the learning curve will cost you money.
4. Can they provide sample traffic data from similar campaigns?
CR benchmarks, approval rates, carrier breakdown. Any legitimate source running MVAS traffic will have this. If they can’t or won’t share it, move on.
5. What’s the minimum test budget and can I cap daily spend?
I never start without a hard daily cap. $50-100 max for day one. If they require a large minimum commitment upfront, that’s a red flag.
The $100 test protocol:
If a source passes the checklist, I run exactly $100 with these parameters:
- Hard daily cap: $50
- Tracking: separate sub-ID for this source only
- Monitoring: check every 4 hours on day one
| Approval rate | 85%+ | 75-85% | Under 75% |
| CR | 6%+ | 3-6% | Under 3% |
| ROI | Positive | -20% to 0% | Under -20% |
[th]
Metric
[/th][th]Green
[/th][th]Yellow
[/th][th]Red
[/th]Yellow on approval rate → pause, investigate source composition
Red on approval rate → cut immediately, don’t optimize
The ongoing monitoring cadence:
Even sources that pass the test degrade over time. My weekly check:
- 3-week rolling average CR (flag if drops 15%+)
- 3-week rolling approval rate (flag if drops 5%+)
- EPC trend (leading indicator — drops before CR does)
Running this process across 5 GEOs. Always looking to expand the source pool with quality in-app and DSP inventory. If you’re running traffic in these markets and want to compare notes on what’s holding quality, I’m genuinely interested in the conversation.
Not here to sell anything. Just think people running the same GEOs should talk more.
