ROI questions deserve numbers, so let me give you the numbers first and the caveats immediately after, because direct mail ROI statistics get repeated online with a confidence the underlying studies do not always support.
The benchmark everyone cites
The most defensible figure comes from the ANA's response rate research: in the widely cited report cycle, letter-size envelope campaigns delivered roughly 112% median ROI, ahead of email at 93% and SMS at around 102%. Compilations like Mailing.com's ROI guide and Postalytics' statistics roundup track these benchmarks across report years, with house lists consistently outperforming prospect lists by multiples.
Read that carefully: 112% ROI means the median campaign roughly doubles its money. It does not mean every campaign does, and the spread around that median is enormous. The same research shows response varying several-fold by list type alone.
The formula, and the inputs that dominate it
ROI = (revenue attributed to the campaign minus campaign cost) divided by campaign cost. Simple arithmetic, three hard inputs. Cost is the easy one; a tracked 6 by 9 postcard typically runs around 80 cents all-in. Attribution is the input most programs fumble, which is why an untracked campaign cannot honestly compute this number at all.
What moves the result most, in order of leverage: the list (mailing people with any prior relationship can multiply response versus cold lists), the offer (a concrete, deadline-bearing offer beats brand prose), and customer value (a $12,000 roofing job forgives a 1% response rate; a $30 pizza order does not). Format and design matter, but they are third-order next to those three.
A worked example
Mail 5,000 postcards at a typical 80-cent all-in rate: $4,000 spent. A 1.5% tracked response is 75 responders. Close a third into customers worth $400 each: 25 customers, $10,000 revenue, and an ROI of about 150%. Now change one input: on a house list responding at 5%, the same math lands 250 responses and ROI goes vertical. That sensitivity is the whole game, and it is why I tell people to obsess over segments, not averages.
My honest expectation-setting: a first cold campaign that breaks even is a success, because you just bought response data by neighborhood. Campaigns two and three, mailed only to the segments that responded, are where direct mail ROI starts embarrassing the benchmarks.