How to find off-market properties: 9 channels ranked by cost per deal

What "off-market" really means

An off-market property is one you can buy without a public MLS listing. No Zillow pin, no yard sign, no bidding war with ten other buyers standing in the driveway. That is the appeal. The problem is that most homes are not off-market. In the National Association of Realtors 2025 Profile of Home Buyers and Sellers, 91% of sellers used an agent and only 5% sold on their own, an all-time low. Most inventory sits behind an agent. Finding the rest is the job.

This guide ranks nine channels by the number that actually pays you: cost per deal. Your results will move with your market, your labor rate, and your follow-up, so treat every figure here as a starting point, verify it against your own market, and consult a professional before you fund a campaign. This is general information, not financial advice.

Cost per lead lies, cost per deal tells the truth

Most guides rank channels by how cheap a lead is. Wrong meter. A free lead that never closes costs more than a $40 lead that does. What matters is the total spend to sign one contract, and that rides on two multipliers almost nobody prints: how many people respond, and how many responses become a deal.

Be honest that those two numbers are assumptions, not measurements. There is no trustworthy published response rate for cold investor mail to strangers. A bogus "3.32%" floated around this business for years and traces back to nothing, so do not build a budget on it. The 2024 ANA Response Rate Report pegs general direct mail around 4% to 9% for house and prospect lists (as reported by Printing Impressions), but that is mail to people who already know you, not cold cards to distressed owners, which pulls far lower. For the math below I use a range: 0.5% to 2% response on one touch, and 10% to 25% of responses signing. Swap in your own once you have them.

Nine channels, ranked by cost per deal

Cheapest per deal at the top. The dollar figures blend sourced unit costs, such as postage and ad clicks and auction deposits, with the labeled response and conversion assumptions above. Read the catch column as closely as the cost.

Driving for dollars: cheapest per deal, hardest to scale

You drive target neighborhoods, log distressed houses, pull the owner, and reach out. The leads are cheap. The miles are slow. Apps like DealMachine bundle the route tracking, owner data, and mail, and start around $99 a month. Turning a plate or address into a phone number runs another $0.02 to $0.15 per record at skip-trace vendors such as DealRun, at roughly 55% to 85% accuracy. So the cash cost of a lead here is close to nothing.

The bill comes due in time, not dollars. One person can only cover so many streets in a Saturday, which is why driving for dollars has the best cost per deal on the table and the worst ceiling. It is the right first channel when you have more hours than money. It is the wrong channel the day you want fifty deals a year instead of five.

Direct mail: the channel that wins the middle

Direct mail looks expensive per lead and reads cheap per deal, which is why it sits in the sweet spot of the table. Run the numbers. A standard First-Class postcard stamp is $0.65 in 2026; all-in with print through a real-estate mail house like Wise Pelican it is about $1.00 to $1.04 a piece. Mail 1,000 pieces and you have spent about $1,000. Not cheap per lead. Cheap per deal.

Two cheaper paths exist, both with limits. USPS Every Door Direct Mail cuts the price to $0.26 a piece, but it blankets an entire carrier route and cannot aim at your list, so it farms an area instead of finding one owner. And that $0.65 card rate only holds at or under 4.25 by 6 inches. Print a jumbo 6-by-9 card and it jumps to letter postage, which quietly wrecks your cost per deal.

At a 1% response that is 10 replies. At 15% of replies signing, that is 1.5 deals, or roughly $670 in mail cost per contract, before your labor. The response rate is an assumption, so treat $670 as the middle of a range, not a promise. Two things make mail work where a single blast fails: targeting the right owners, and hitting them more than once. This is where direct mail stops being a numbers game and starts being a list game.

Farmrix is built for that list game. It scores every owner in a market on how likely they are to sell in the next 6 to 12 months, ranks them, and mails postcards to the top of the list instead of the whole county. Fewer pieces, aimed at the owners most likely to actually deal. That is how a mid-table cost per deal turns into a channel you can run every month.

Public-records lists: foreclosure, probate, absentee, vacant

This is direct mail's ammunition, and most of it is public and cheap. County and court records surface owners under real pressure: pre-foreclosure, probate, code violations, long-time absentee landlords. The universe is bigger than people think. ATTOM counted 227,548 U.S. properties with foreclosure filings in the first half of 2026, up 21% year over year, though that is still just one in every 632 housing units, so this is a targeted signal, not a flood.

Look underneath that headline. ATTOM logged 164,566 foreclosure starts, up 18%, plus 27,983 completed repossessions in the same half of 2026. Of the 245,376 homes in the foreclosure process in Q2, only 8,312 sat vacant as "zombie" foreclosures. The genuinely abandoned pool is smaller, and more findable, than the scare headlines imply.

Vacant and absentee lists are the quiet winners. ATTOM's Q2 2026 report found nearly 1.4 million vacant homes, and investor-owned homes were vacant at 3.5% versus 1.3% overall, nearly three times the rate. A tired landlord with an empty rental is a motivated seller who has not told anyone yet. Pull the list, skip-trace it, and mail it before the foreclosure feeding frenzy ever starts.

Raw county lists are messy: duplicates, dead addresses, owners who sold last year. Scrubbing and cross-referencing all of it by hand is the real tax on this channel. Farmrix does that cleanup and hands back a ranked, mailable list, which is why public records and direct mail work best as one move instead of two.

Wholesalers and auctions: fast, but you pay for it

Two channels trade cost for speed. Wholesalers hand you a deal already under contract and assign it to you for a fee, often several thousand dollars baked into the price. You close fast, but the discount that made it a deal is now partly the wholesaler's margin, not yours. Fine for volume buyers, expensive per deal.

Foreclosure auctions are the capital-heavy end. Rules vary by county, and they are strict. St. Johns County, Florida requires a 5% deposit of your high bid on the spot, with the full balance due by 2:00 PM the next business day or you forfeit the deposit. No financing, no inspection, no warranty, and you are bidding against people who do this full-time. Great for deep-pocket buyers, brutal for beginners.

There is a second risk beyond the cash. You usually buy with no title search and no walk-through, and depending on lien position some debts survive the sale, so a winning bid can hide a surviving mortgage or a gutted interior. Read the county rules and the specific case file before you raise your hand.

Agents, pocket listings, and the PPC trap

Agents control most of the market, so relationships open doors, especially to pocket listings that never hit the MLS. Those are real but limited. A 2021 Redfin survey found 43% of agents said pocket listings had grown, yet only 2.3% of homes were marked sold or pending the same day they listed, and the NAR Clear Cooperation Policy now pushes most listings onto the MLS within a day. Redfin put about 21.4% of homes as sold off-MLS in that period, so the channel is real but a minority. Networking is low cash and slow, and worth doing anyway.

Then there is the advice to "just run Google Ads for we buy houses." Watch the math kill it. In July 2026 the average cost per click on "we buy houses" was $110.73, and "sell my house fast" ran $133.23 (DataForSEO). Say your landing page converts a strong 5% of clicks into a real lead. That is 20 clicks, about $2,215, for one lead. If 15% of leads sign, you need close to seven leads per deal, roughly $15,000 in ad spend for a single contract before you pay yourself a cent. That is not a channel. It is a bonfire. Direct mail at that same 15% conversion came in near $670, more than twenty times cheaper for the identical funnel math. PPC earns its place only once you have the volume and brand to convert far better than 5%.

How to choose your channel

Match the channel to what you actually have. More time than money: drive for dollars and work public records by hand. More money than time: mail a targeted list and hit it monthly. Deep pockets and nerve: stand at the auction. Skip PPC until your conversion is proven, and treat wholesalers as a supplement, not a strategy.

For most investors the honest answer is the middle of the table. A ranked list plus repeat direct mail gives you the best mix of cost, control, and scale, which is exactly what Farmrix does: it finds the owners most likely to sell and mails them for you, starting at 500 ranked owners and 500 postcards for $1,195. Pick one channel. Run it ninety days. Count deals, not leads. Less mail, more deals.