Six Numbers Per Record, and Nobody to Dial Them

The pitch for skip tracing is that it solves your lead problem. It does not. It relocates it.

Here is the arithmetic nobody runs before they buy. Per Wholster, skip tracing costs about $0.12 per record in the 1,000 to 10,000 range and drops toward $0.10 at volume, with 85 to 89% accuracy and roughly a 90% match rate on delivered records. Each record can return up to six phone numbers, plus emails and mailing addresses. Lists come in more than thirty flavors: absentee owners, vacant properties, tired landlords, probate, preforeclosure, distressed. Order in the morning and the file is dial-ready in 24 to 48 hours.

So you spend $500. You get back roughly 5,000 records. And roughly 30,000 phone numbers.

The moment the file lands

Now watch what actually happens in the office.

Somebody opens the file. Somebody sorts it. Somebody decides which of the six numbers to try first, which is a guess, because nothing in the file tells you which line the owner actually answers. Somebody starts dialing at 9am and by 11:30 has made maybe eighty calls and reached four humans, two of whom were the wrong person and one of whom was extremely clear about it.

By Thursday the file is "in progress." By the following Thursday it is a tab nobody clicks. Three weeks later somebody buys a fresh list, because the old one feels stale, and the cycle restarts. The data was fine. The throughput was never there.

The gap is capacity, not quality

This is the honest read on skip tracing. The vendors are not overselling accuracy. 85 to 89% is a real number and it is good. The problem is that a good number multiplied by six, multiplied by thousands of records, produces a work queue that has no relationship to how many hours a human calling team actually has.

Cold outbound already runs on brutal ratios. Reaching an appointment takes something on the order of 330 dials in this market, a figure we have written about before. Now stack that against 30,000 numbers and ask how many appointments are theoretically sitting in the file, then ask how many your team will ever get to.

The answer is usually the same: a single-digit percentage of the list, worked once, in the first ten days, and never again.

Meanwhile Wholster notes that 10 to 15% of returned numbers are bad. Every one of those is a dial your team spends to learn nothing. Nobody budgets for that. It shows up as a bad week and a demoralized caller.

What machines are actually good for here

The useful framing is not "AI replaces the caller." It is that the file has three distinct jobs in it, and only one of them needs a person.

Validation is job one. Dedupe across records, strip the disconnected lines, rank the six numbers by likelihood of pickup instead of by column order. That is pure computation and it recovers the 10 to 15% waste before a human hears a single ring.

Volume is job two. Working thirty thousand numbers on a schedule, across time zones, with retries spaced sensibly rather than in a panic on day three, is not a thing a five-person team does. It is a thing a system does while the team sleeps.

Conversation is job three, and that is where the humans belong. A live seller who says the word "probate" out loud deserves a person, immediately, with the whole context in front of them. That conversation is worth more than the entire day of dialing that produced it.

Split the file that way and the economics stop being absurd. Your $500 of data stops being an aspiration and starts being a queue that actually gets worked.

The uncomfortable question

If you bought a list in the last ninety days, go look at it. Not the summary. The actual file.

Count how many records were never dialed once. That number is what you paid for and did not use, and it is almost always larger than the number people expect. It is also the cheapest pipeline you will ever find, because you already own it.

The data was never the bottleneck. The dial tone was.

Track A note: soft CTA below for the WholesaleAI version. Strip for any non-business placement.

An AI Operations Audit counts what's actually in your lists, what got worked, and what a system could reach that your team never will. Book one before you buy the next file.