If your program is mobile-heavy in South Africa, some conversions your partners drove are being mis-credited to “direct” or dropped entirely — a CGNAT problem. Here’s a practical way to measure the loss and recover it, without ripping out your stack.
Step 1 — Confirm you actually have the problem
You probably do if two things are true: your traffic is majority mobile, and a meaningful slice of conversions land as “direct / unattributed.” That bucket is where mis-resolved mobile conversions hide. The bigger it is relative to your mobile share, the more you’re likely mispaying partners and under-crediting channels.
Step 2 — Measure it on your own traffic (the audit)
Don’t optimise on a guess. Run a two-week Attribution Audit: place one tracking link on your mobile flow, then compare what a CGNAT-aware method resolves against what your current attribution reports. The output is a single number — how much attribution you’re missing — with your data, not a vendor benchmark. If the gap is small, you’ve lost nothing but two weeks. If it’s large, you’ve found money.
Step 3 — Recover with a device-bound match
The recovery mechanism is to anchor each click to a cryptographic, device-bound token rather than the rotating IP. When a conversion arrives — even from a different IP, even with no surviving cookie — the token resolves it back to the partner who earned it. Credit lands correctly; your best mobile channels stop looking artificially weak.
Step 4 — Pay on what’s verified
Recovery is only useful if it’s trustworthy, so the recovered conversions should be verifiable and reconciled to payouts — FICA-gated, POPIA-compliant, no personal data stored to make the match. That’s what lets your finance team sign off on paying partners for conversions that legacy tracking missed.
What good looks like
- The “direct/unattributed” bucket shrinks toward its true size.
- Mobile channels show their real CPA/ROAS — usually better than you thought.
- Partners are paid for what they actually drove, which strengthens the relationships that scale you.
The honest way to start is to measure, not to trust a claim. If you want the gap quantified on your traffic, request a demo and we’ll scope a two-week audit.