Every pay-per-call buyer has taken this call. The number looks local. The caller asks two plausible questions, stays on the line past the payable threshold, then evaporates. No sale, no callback, no voicemail box. Multiply it by forty a week and it stops being noise and starts being a line item.
Most of those calls come from burner apps. Here is how one actually gets made, and where it leaves fingerprints.
Where burner numbers come from
TextNow, Google Voice, Talkatone, Hushed, and a hundred smaller apps all sell the same thing: a working US phone number in about ninety seconds, no identity attached. Under the hood they draw from pools of non-fixed VoIP numbers, lines that exist entirely in software and are not anchored to a physical address or a carrier account with a name on it.
That is the entire product. For a traveling contractor or someone selling a couch online, it is a genuinely useful one. For someone paid per call, it is a fraud kit with a free tier.
The lifecycle of a burner campaign
A burner operation is boring and repetitive, which is exactly why it scales:
- Spin up. Register a batch of numbers across a few apps. Pick area codes that match the campaign's geo targeting. This takes minutes and costs nothing.
- Dial. Work through the promo numbers, one campaign at a time. Incentivized traffic, pay-per-call affiliate offers, and lead-gen campaigns with duration-based payouts are the usual targets.
- Hold. Stay on the line past the billable threshold. Scripts help. Some operations run soundboards; the lazier ones just mumble through insurance questions for ninety seconds.
- Burn. Once a number starts getting flagged or blocked, release it and pull a fresh one from the pool. The app does not care. The pool is deep.
Notice what never appears in that lifecycle: a real prospect.
Why your eyeball checks miss them
The classic manual defenses all fail against this pattern. The area code is right, because the fraudster picked it. The number is not on any shared blocklist yet, because it is four hours old. The call duration looks healthy, because holding the line is the whole job. By the time a human notices the pattern, the number is already burned and replaced.
Blocklists chase yesterday's numbers. Burner pools mint tomorrow's.
The fingerprint that survives
One attribute does not rotate with the number: the line type. A burner app number is non-fixed VoIP on the day it is created and non-fixed VoIP on the day it is burned. Carrier-grade intelligence sees that classification the moment the number is looked up, whether the number is four years old or four minutes.
That is the core of how CallerSift screens. When your tracker pings your lookup URL, the verdict weighs what the number is, not just whether anyone has complained about it yet. A fresh burner earns a block on its very first call, which is precisely the call a blocklist can never catch.
What real traffic looks like
Roughly speaking, genuine consumer calls skew heavily toward mobile lines, with landlines behind them. Real people occasionally call from VoIP lines too, which is why a sane screening setup treats line type as strong signal rather than a hanging judge, and why CallerSift lets you review every verdict, reason attached, in the lookups view.
What to do about it
If you buy or broker calls, three habits close most of the gap:
- Screen before the route. A burner call that never rings a buyer costs you nothing and teaches the operator your campaign is a dead end.
- Watch block rate by source. Burner traffic clusters. One affiliate at a 30 percent block rate is not unlucky, they are supplying manufactured calls.
- Stop paying for held air. Duration thresholds alone reward exactly the behavior burner operations are built around.
CallerSift does the first one in milliseconds and hands you the data for the other two. Start screening and watch what your traffic is actually made of.