The list is the treasure map. Bad map equals bad ROI.
Before any of the math below — internalize this. A great letter to a bad list flops. A mediocre letter to a great list works. The list is the treasure map; the mail is the shovel. A sharp shovel digging in an empty field still finds nothing. Operators who spend their effort on copy tweaks while running a stale or generic list are sharpening the shovel and ignoring the map.
Everything else in this article — response rate, cost-per-call, ROI — those metrics tell you whether the map was good or bad. Use them. But the upstream decision (which list, from which vendor, refreshed how often) is what makes those numbers good or bad in the first place.
“What’s a good response rate?” is the wrong question.
A 4% response rate sounds great until you find out the campaign cost $8,000 and produced zero contracts.
A 0.6% response rate sounds bad until you find out the campaign cost $1,500 and landed one $22,000 wholesale assignment.
The number that pays your mortgage isn’t response rate — it’s cost-per-deal. Response rate is an intermediate metric, useful for diagnosing whether a campaign is healthy mid-run, but useless as the bottom-line measure.
Forty years of mailing for real estate investors and the operators who survive are the ones who measure money in, money out. The ones who flame out chase response rate because it feels concrete — and they end up with a high response rate on a list of tire-kickers who never close.
The three numbers, ranked by what matters.
1. Cost-per-deal (CPD) — the only number that matters
Formula: Total campaign cost ÷ contracts closed (or, for buy-and-hold, properties acquired).
Example: $2,940 campaign (2,000 pieces × $1.47) lands 3 contracts. CPD = $980.
This is the number to optimize. Lower CPD means more deals per marketing dollar. Period.
A campaign with 0.4% response and 1 closed deal at $1,500 spend has a CPD of $1,500. A campaign with 3% response and 0 closed deals has a CPD of infinity.
The first one is a winner. The second is bankruptcy.
2. Cost-per-call (CPC) — the diagnostic
Formula: Total campaign cost ÷ inbound calls.
This is the leading indicator. You’ll know CPC weeks before you know CPD (because calls come in, then contracts come slower). If your CPC is in line with historical norms for your market, the campaign is healthy. If it’s wildly higher, the list is bad, the letter is bad, or your phone number is broken.
Typical real estate CPC ranges: $50-200 per call. Below $50 means you’re getting tire-kickers or the list is hot. Above $200 means something’s off and the campaign won’t produce deals at acceptable CPD.
3. Response rate — the easy-to-game number
Formula: Inbound calls ÷ pieces mailed × 100.
This is the metric direct-mail “gurus” love to quote because it’s easy to inflate and easy to compare. It’s also the most misleading.
A 5% response rate from “anyone with a heartbeat who picks up the phone” is worse than a 0.5% response rate of qualified motivated sellers who actually own the property.
Use response rate to compare campaigns to each other — not as a target.
Industry baseline numbers (and why they lie).
The Data & Marketing Association publishes annual response-rate benchmarks that get cited everywhere. They claim direct mail averages 4-9% response. Real estate direct mail “gurus” cite those numbers, then their students try to hit them and feel like failures when they pull 1%.
Here’s the honest baseline for real estate direct mail to motivated sellers. Read this carefully — response rates vary wildly by market. Plan around the typical baseline; treat the exceptional high as upside if your specific market delivers it, not as the number to model your campaign math on.
| List type | Typical baseline | Exceptional high (specific markets/openers) |
|---|---|---|
| Generic absentee owner | 0.3–0.8% | up to 1.2% |
| Skip-traced absentee with equity | 0.4–1% | up to 1.5% |
| High-equity + long tenure | 0.5–1.2% | up to 2% |
| Pre-foreclosure (fresh) | 0.8–2% | up to 3% on weekly-fresh in specific markets |
| Probate (fresh, weekly pulls) | 1–1.5% | up to 3% in less-saturated markets |
| Pre-probate (death certificates) | 0.8–2% | up to 5–8% on the “I have a question” curiosity opener in some inheritance markets |
These come from watching our customers’ campaigns for 40 years. The single biggest variable is the market itself — same list type, same letter copy, in two metros, produces dramatically different numbers based on local competition density. The 8% pre-probate case at the top is a real number we’ve seen, but it’s opener-specific (the “I have a question about your property” curiosity mailer), list-specific (fresh inheritance), and market-specific (low-competition metro). Most operators on most lists pull the typical baseline.
If you’re consistently above the typical baseline, your market or your opener is doing extra work for you. Enjoy it. Don’t bet next year’s plan on it continuing.
If you’re consistently below the typical baseline, the answer is usually the list (stale, generic, oversold). Sometimes it’s the copy (predatory, branded, too long). Almost never is it “direct mail doesn’t work.”
How to instrument a campaign for honest tracking.
You can’t manage what you don’t measure. Most operators measure nothing, then declare direct mail dead when they can’t tell which campaign worked.
The minimum-viable tracking stack:
One unique phone number per campaign
Tools: Google Voice (free), OpenPhone ($15/mo), CallRail ($45/mo for full tracking).
Use a different number on each campaign — even different numbers per touch within the same campaign. Every inbound call instantly tells you which mailing generated it. This is the single highest-ROI piece of infrastructure you can build into your direct mail.
A simple call log
Spreadsheet works fine. Columns:
- Date of call
- Phone number called (= which campaign)
- Caller name + property address
- Disposition (interested / not interested / wrong number / hang-up)
- Notes
You’ll know within 4 weeks which lists are pulling and which aren’t.
A deal log
Same spreadsheet, second tab:
- Date of contract
- Source campaign (linked to call log)
- Deal type (wholesale assignment / fix-and-flip / hold)
- Margin (or projected margin)
Now you can calculate CPD per campaign with a single formula.
What NOT to track (yet)
Skip the fancy CRM with 47 fields per lead. Skip the lead-scoring algorithm. Skip the multi-touch attribution model. Those are problems for operators doing 200+ deals a year. If you’re doing 1-20 deals a year, the spreadsheet stack above tells you everything you need.
The ROI math operators actually need.
Three calculations. Memorize them.
Per-campaign ROI
ROI = (Deals × Avg Margin - Campaign Cost) / Campaign CostExample: 1 wholesale assignment at $18,000 margin on a $1,470 campaign. ROI = ($18,000 - $1,470) / $1,470 = 11.2× return
Maximum acceptable cost-per-deal
Before you start a campaign, decide: how much would I spend per deal and still consider this a win?
For a wholesaler with $15k average margin, paying $3k per deal is a 5× return — that’s a good business. Paying $7k per deal is a 2× return — marginal. Paying $12k per deal is upside-down.
Set your ceiling before you start. When CPD exceeds the ceiling, kill the campaign.
Break-even response rate
Break-even response rate = (Campaign cost / Avg margin / Conversion-call-to-deal) × 100Example: $1.47 per piece, $15k avg margin, 5% call-to-deal conversion. = ($1.47 / $15,000 / 0.05) × 100 = 0.196%
You need 0.2% response just to break even. Anything above that is profit.
Knowing this number ahead of time tells you what bar a campaign has to clear. Most real estate direct mail beats 0.2% comfortably — but if you’re selling something with thinner margins or worse call conversion, the break-even rate shifts up fast.
Why touch 2 changes everything.
The most common measurement mistake we see: judging a campaign after touch 1.
A single-touch campaign at 0.6% response looks marginal. The same list with a second touch five weeks later typically adds another 0.5-0.8% — bringing the cumulative response to 1.1-1.4%. The campaign cost increased 100% (you mailed twice), but the response increased 80-130% — improving CPD significantly.
Touch 2 isn’t a fresh attempt. It’s the same names you already mailed, who saw your first letter, set it aside, and are now in a slightly different mental state. Some had a life event in the intervening five weeks. Some thought about it. Some lost the first letter and now want to call.
Skipping touch 2 because touch 1 “underperformed” is the most expensive mistake in real estate direct mail.
Cost-per-lead vs cost-per-deal vs cost-per-piece.
Three numbers, three different uses:
- Cost-per-piece ($1.47–1.52 with Yellow Letter): What you pay to put one envelope in one mailbox. This is the input cost.
- Cost-per-lead (= cost-per-call, $50–200 typical): What you pay per inbound conversation. This is the leading indicator.
- Cost-per-deal ($1,000–5,000 typical): What you pay per closed transaction. This is the bottom line.
A campaign can have a great cost-per-piece (you got a bulk discount), a mediocre cost-per-call (response was lower than expected), and a great cost-per-deal (the few calls that came in closed at unusually high rates). That’s a winning campaign — even though one of the intermediate metrics looked soft.
Don’t confuse the layers.
When to kill a campaign.
Two conditions, both required:
- Cost-per-call exceeds your acceptable maximum. If your model needs sub-$100 CPC and you’re at $250 after 60 days, the list isn’t generating enough volume.
- You’re at least 90 days past the last touch. Real estate direct mail produces a long tail of late responders. Killing at 30 days throws away normal late calls.
Both conditions matter. Killing on time alone (without checking CPC) leaves you guessing. Killing on CPC alone (without giving it 90 days) cuts off legit response.
Common measurement mistakes.
- Not using unique phone numbers. “How did you hear about us?” is unreliable — half the time the caller says “I don’t remember” or names a campaign you didn’t run.
- Counting wrong-number calls as responses. Phone validation on your list is a real expense, but it pays for itself by cleaning the noise out of your response metric.
- Comparing apples to oranges. A 1% response rate on a probate list is poor; a 1% response rate on a generic absentee list is good. Compare a list type to itself across campaigns, not to other list types.
- Ignoring the time dimension. A campaign that pulled 0.4% in week 1 and 0.6% in week 6 is converging on 1% by week 12. The early window underrates the campaign.
- Optimizing for the wrong metric. A copy change that improves response rate but tanks call quality (more tire-kickers, fewer real sellers) is making the campaign worse, not better.
Frequently asked.
What’s a good direct mail response rate? 0.5-2% is normal for real estate motivated-seller mail. 2-5% is excellent (usually fresh probate or pre-foreclosure). Anything claiming 10%+ is hype or a different metric. The number that matters more is cost-per-deal.
How do I track direct mail responses? Unique phone number per campaign (Google Voice, OpenPhone, CallRail). Every inbound call is attributable to the campaign that generated it.
What’s the difference between response rate, conversion rate, and ROI? Response rate = calls ÷ pieces. Conversion rate = contracts ÷ calls. ROI = (deals × margin) ÷ campaign cost. They measure different things. ROI is the only one that pays your bills.
How long should I wait before judging a campaign? 90 days minimum after the last touch drops. Real estate direct mail produces late responses for weeks after the mail lands.
Should I A/B test my direct mail? Only with meaningful volume (1,000+ per variant). Below that, randomness will look like a winner. Most small operators are better off improving list quality.
Why does my response rate vary so much between campaigns? List quality, almost always. 70% list-quality, 30% copy/format.
What’s a typical cost-per-deal? $1,500-5,000 for motivated-seller mail. A $4,000 campaign that lands one $25,000 contract is a winner.
When should I kill a campaign? When cost-per-call exceeds your ceiling AND you’re 90+ days past final touch. Both conditions required.