Data
The hidden cost of bad lead data
Every agent has had this moment: you sit down for your dial window with a list of 40 leads. By the third call you've gotten a wrong number, a disconnected line, and a guy who sold the property in 2018 and is "tired of being called about it." By call number ten you're not just frustrated — you're convinced you can't sell.
Bad lead data does that. It taxes you twice: once when you dial it, and again when it tanks the only psychological resource that matters in cold outreach — your willingness to pick up the phone for the next call.
Most agents underestimate the cost. They think a 60% match rate is "fine" because it's a number, and numbers feel objective. They don't run the math on what those bad rows actually cost them in dollars, time, and morale. Let's run it.
Where bad data comes from
Before we cost it, it helps to know what we're actually dealing with. In a typical real estate prospecting list, "bad data" is some combination of:
- Wrong owner of record — sold years ago, county records still stale, or the listing system mis-merged a row
- Owner is an LLC or trust — the entity owns the property but the person you need to talk to is two layers of corporate paperwork away
- Bad phone number — disconnected, reassigned to a stranger, or a landline that hasn't worked since 2014
- Out-of-state mailing address — owner lives across the country, has no intention of selling the rental, doesn't take cold calls
- Co-owners not surfaced — you call John but Mary is the actual decision-maker and she's never on your list
- DNC list / known litigators — calling them is at best a waste, at worst a $500-1500 TCPA exposure per call
Each of these has a real, calculable cost. Here's what a typical week of bad data looks like for a working agent.
The real per-week cost
Assume an agent who runs 200 prospects per week with a (very common) 65% usable contact rate. That's 70 bad rows per week. Here's where the time actually goes:
| Bad row category | Time per row | Weekly cost |
|---|---|---|
| Wrong number / disconnected (~25 rows) | ~90 sec | ~38 min |
| Wrong owner of record (~15 rows) | ~3 min | ~45 min |
| LLC / trust unsurfaced (~10 rows) | ~5 min | ~50 min |
| Out-of-state mailing (~12 rows) | ~2 min | ~24 min |
| Manual scrub / "let me Google this" (~8 rows) | ~6 min | ~48 min |
That's a little over 3.5 hours per week just on dealing with bad rows — not counting the morale hit, which is the more expensive cost and the harder one to measure.
The morale tax. The Princeton "default mode network" research is well established: cognitive friction in the first ten minutes of a task drops total task output by 30-40%. Translation: if your first five calls are wrong numbers, you make fewer total dials that day — regardless of how many good rows are sitting in the queue below.
The TCPA layer most agents ignore
Bad data isn't just inefficient. Calling a number that's on the federal Do Not Call list, or worse, a known TCPA litigator, can cost $500 to $1,500 per call in statutory damages. There are professional plaintiffs whose entire business model is logging real-estate cold calls and sending demand letters.
A list with no DNC scrub and no litigator flag is not a "60% usable" list. It's a 60% usable list with a tail-risk lottery ticket attached. One bad call to the wrong person can erase a year of commission.
The only correct way to handle this: every list, every time, gets scrubbed against the DNC registry and litigator database before you dial. The flag should appear on the lead card. If it's flagged, you skip — or you call with documented prior express consent. There is no third option.
What "good data" actually looks like
A clean list, ready to dial, has these properties on every row:
- Verified owner of record, sourced from county parcel data within the last 30 days
- Both co-owners surfaced when joint ownership is on title
- LLC / trust unmasked where possible — registered agent and managing member resolved to a real human
- Phone numbers ranked by recency, not just dumped in a list — the most-likely-to-work number ringfirst
- Emails when available, so you have a second channel for the no-answers
- DNC + TCPA litigator flags visible at a glance
- Equity / ownership-length signals — so you can prioritize who to actually call
Anything less and you're paying twice for the same lead: once to source it, and again in the cognitive cost of triaging it on the phone.
The "no data, no charge" rule
Here's the rule we wish every prospecting tool adopted: if a row comes back without usable contact data, it doesn't cost a credit.
Most vendors charge per attempt. You upload 100 rows, they bill 100 credits, and 35 of them come back empty. You just paid for nothing — and the 35 missed rows are the worst of both worlds, because you have neither the contact data nor the budget to look elsewhere.
The only honest pricing model is: pay for the matches. Misses don't cost a credit. Same owner across multiple properties counts as one charge. Anything else and the vendor's incentives are pointed away from accuracy and toward churn.
What to do this week
Three concrete steps:
- Run the audit. For one week, log every "bad row" you hit. Categorize them. The number will be bigger than you expect.
- Demand the flags. If your current data source doesn't surface DNC, litigator, and second-owner data on every row, switch. This is non-negotiable in 2026.
- Switch to a "matches only" billing model. If you're paying for misses, you're funding bad data. Stop doing that.
The agents who treat data quality as an operational discipline — not as a vendor problem to outsource and forget — close more deals with less burnout. That's the entire game.