Direct mail advertising: the arithmetic that decides whether to send it, and the four decisions before you print
Saturation by route against a targeted list, every cost line so quotes are comparable, the break even calculation in five minutes, and the four ways to measure something that arrives on paper.
Published by AI Knows Us · Updated 19 October 2026
Direct mail is the one advertising channel that arrives in somebody's hand and stays there until they throw it away, which is why it still works in categories where email is ignored and social is skipped. It is bought two ways and the choice decides everything else. Saturation mail goes to every address on chosen postal routes, cheaply, with no list to buy, and it suits a business whose customers are simply whoever lives nearby. Targeted mail goes to a named list, costs considerably more per piece, and suits a business that can describe its buyer precisely. The arithmetic is unforgiving in a useful way: you know the cost per thousand delivered before you print, so you can calculate the response rate you need to break even and decide in advance whether that number is plausible for your offer.
This page covers the two routes, the four decisions before you print, every cost line, the break even arithmetic and the four ways to measure something that arrives on paper. For the production specification and how to brief a printer, see marketing material printing.
The two ways to buy it
| Saturation by route | A targeted list | |
|---|---|---|
| Who gets it | Every address, or every residential address, on the postal routes you select | Named households or businesses matching criteria you set |
| How you buy it | Through the postal service's own saturation product, in the United States the Every Door Direct Mail service, selecting carrier routes on a map | A list rented from a broker or compiled from your own data, then mailed through a mailing house |
| Cost per piece | Lowest available, because there is no list cost and the postage rate is designed for volume by route | Higher, sometimes several times higher once the list, the addressing and the handling are counted |
| Best for | Local trades, restaurants, retail, anything where the catchment is the audience | High value purchases, business to business, and any offer that only makes sense to a narrow group |
| The trap | Paying to reach thousands of people who could never buy from you, which is fine at low cost per piece and ruinous at high | List quality. An old list produces returns, waste and a poor response you will wrongly blame on the creative |
Your own customer and enquiry list sits outside both and beats both. Mailing people who have already bought from you is the highest responding mail most businesses will ever send, and it costs only postage and print.
The four decisions before anything is printed
- Who gets it. Routes or list, and how many. Decide the quantity from what you can handle if it works, not from the price break. A plumber who mails ten thousand homes and can serve forty jobs a week has bought a problem.
- The format. A postcard is cheapest, needs no opening and gives you two sides. A letter in an envelope costs more, gets opened less and carries a longer argument when it is opened. A larger flat piece commands attention and costs more in postage. Check the postal size and weight categories before the designer starts, because a few millimetres can move a mailing into a higher rate.
- The offer. This decides the result more than the design does. It needs to be specific, time bound and easy to redeem, and it needs to be worth the trip. A vague invitation to visit the website is not an offer.
- How you will know. Decided before printing, because every method below requires something printed on the piece.
Every cost line, so the quotes are comparable
- The list, or nothing for saturation routes. Rented lists are usually priced per thousand names, often for one use only.
- Design. One piece, two sides, and the variants if you are testing.
- Print. Per thousand, falling steeply with volume, and dependent on the stock and the finish.
- Postage. Normally the largest line. Rates depend on size, weight, shape and whether the mail is pre sorted, and the saturation products exist specifically to lower this.
- Mailing house handling. Addressing, sorting, inserting, bundling and delivery to the postal depot. Small jobs can be done by hand and large ones cannot.
- The redemption. If the offer is a discount, that discount is an advertising cost. Count it.
The break even arithmetic, which takes five minutes
Direct mail is the rare channel where you can do this before committing, because the cost is known in advance.
- Work out the total cost of the drop from the lines above, and divide by the number of pieces. That is your cost per piece delivered.
- Work out the gross margin on one sale, not the revenue. For a service with repeat visits, use the margin on the first job plus a conservative view of the second.
- Divide the cost per piece by the margin per sale. That gives the response rate you need to break even, usually expressed as a fraction of one per cent.
- Ask whether that number is plausible for this offer and this audience. Not against an industry average, which varies so widely by category and list quality that it tells you nothing about your drop. Against your own judgement of the offer, and against your own result once you have one.
- Then decide the test size. Mail enough to produce a readable number of responses rather than a statistically meaningless handful, and hold back the rest of the list or the adjacent routes for the second drop.
The first drop buys you the one figure that matters, which is your own response rate. Every drop after that is a business decision rather than a gamble.
Four ways to measure paper
- A code that appears nowhere else. A discount code, a reference, a promotion name. Simple, and it undercounts, because plenty of people respond without quoting it.
- A dedicated web address. A short, memorable path used only on this mailing, pointing at a landing page that matches the offer. Visits to it are a clean count.
- A separate phone number. For any business where people phone, this is the strongest single method, and it captures the response that no code or URL ever sees.
- A matched market holdout. Mail some routes or postcodes and deliberately not others of similar profile, then compare total enquiries between the two. This is the only method that measures what the mailing caused rather than what it was credited with, and it is the one to use when you are deciding whether to keep doing it.
Also match the addresses you mailed against the addresses that ordered, if your system records them. For a saturation drop over a defined area, the simplest signal of all is the change in enquiries from that area in the three weeks after delivery.
Five mistakes that account for most failed mailings
- No offer. A piece that announces the business rather than asking for something specific.
- One drop. Recognition builds over repeated contact with the same households. A single mailing to a cold list is the most expensive way to buy a small result.
- A list nobody cleaned. Returns are wasted postage and print, and a high return rate is evidence the list was sold to you tired.
- No way to measure, decided afterwards. By then the piece is printed and the method had to be on it.
- Mailing at the wrong moment for the category. Roofing in the week after a storm, tax services before the deadline, restaurants on a Thursday. Timing is a bigger lever in mail than in most channels, because the piece sits on a counter for days.
The rules to keep in mind
Advertising claims on paper are held to the same standard as anywhere else: a price must be the price, a discount must have genuine terms printed, and an expiry must be real. Rented lists come with contractual limits on how many times you may use them and on what you may do with the names afterwards, and the contract is worth reading rather than assuming. If you are building your own mailing list from customer records, the consent and data rules of your market apply to postal addresses as much as to email, and they differ by country, so settle that before the first drop rather than after a complaint.
Direct mail advertising: what people ask
Does direct mail still work?
In several categories it works better than it did, for a straightforward reason: the inbox is crowded and the letterbox is not. It performs most reliably for local services with a defined catchment, for high value considered purchases where a physical piece carries weight, and for mailing your own past customers. It performs worst as a substitute for having an offer.
What response rate should I expect?
Published averages span such a wide range across categories, list types and offers that planning against one is a mistake. Do it the other way round: calculate the response rate you need to break even from your own costs and margin, judge whether that is plausible, then measure your actual rate on a test drop and use that as your benchmark from then on.
How much does a mailing cost?
Price it per thousand delivered rather than per piece, and get quotes on all the lines above separately, because print and postage move independently. The cheapest route by a wide margin is saturation mail by postal route, where there is no list cost and the postage product is built for volume. The most expensive is a small run to a rented list with a bespoke format.
Postcard or letter?
Postcard for a simple local offer that has to be understood in two seconds on the way in from the door. Letter when the argument needs more than a few sentences, the purchase is considered, and the recipient has a reason to open something addressed to them personally. Test one against the other on the same list before committing a year's budget to either.
Can I mail my own customer list?
Yes, and it is usually the best mail you will send. Check the consent and data protection position for postal contact in your market first, keep an easy way for people to ask you to stop, and honour it immediately.
How many times should the same household get it?
More than once. A sequence of two or three pieces over a season to the same routes, with the same look and a consistent offer, reliably outperforms the same budget spread across three times as many addresses once.
The piece that is still on the counter
A mailing has a long tail that no other channel has: it sits in a kitchen for a fortnight, and when the need finally arrives somebody picks it up and looks the business up. That lookup is where the mailing either converts or quietly fails, and increasingly it is a question put to an assistant rather than a search. A piece of paper with your name on it creates the intent, and something else entirely decides what the buyer is told about you at the moment they act on it. It is worth knowing what that answer currently says before you print ten thousand of anything. Measuring whether assistants name you sets out how to take that reading.