Been working on an eSIM platform called Mesim for a while now and honestly one of the hardest parts hasn’t even been building the product…
It’s been dealing with fraud.
At the start I was running Meta ads and thought things were going crazy in a good way. Cheap conversions, heaps of signups, traffic looked solid etc.
Then a few weeks later:
- chargebacks everywhere
- stolen cards
- fake names
- people smashing data instantly then disappearing
- random countries I wasn’t even targeting
Because eSIMs are digital and activate instantly, fraudsters absolutely love them.
You don’t really get a chance to stop it once the order goes through.
I tried a bunch of the standard fraud tools but most seem more built for physical ecommerce stores. eSIM is different because there’s no shipping delay or anything to verify against.
Ended up building a heap of custom stuff into the checkout flow:
- Stripe Identity verification
- VPN/proxy detection
- disposable email blocking
- risk scoring
- country mismatch checks
- velocity limits
- manual review on certain orders
Biggest thing that helped was forcing some users through ID verification before they could activate.
Conversion rate dropped a bit but the quality of customers became way better and chargebacks dropped hard.
Honestly feels like nobody talks about this side of SaaS enough. Everyone posts revenue screenshots but not the part where fraud can completely destroy ad campaigns.
Curious how other SaaS founders handle this especially if you’re selling digital products instantly.
Also if anyone has good recommendations around:
- fraud prevention
- scaling Meta ads
- reducing chargebacks
- international traffic quality
keen to hear them.
Top comments (1)
This is a massive reality check for anyone scaling digital products. The 'instant activation' feature of eSIMs is a double-edged sword that fraudsters clearly love. At Aqva Marketing, we often see Meta ads bringing in high volume, but if the pixel isn't optimized for 'Quality Lead' events rather than just 'Signups,' it can feed the algorithm the wrong data. Have you tried setting up Offline Conversions to feed the fraud/chargeback data back to Meta? It can sometimes help the AI stop targeting those low-quality segments