Circulation is not customers: how to measure what print advertising earns

Print can be measured publication by publication. Give each one a door no other channel uses — its own number, a print-only page, a named answer to "How did you hear about us?" — and match those leads to closed sales for a floor. Circulation counts copies, not customers; readers who search your name need a zoned or on-off comparison.

A kitchen and bath remodeler runs ads in three places: a half page in the metro daily's Sunday home section, a quarter page every week in the community paper, and a full page in a regional lifestyle magazine. The rate cards quote circulation, the sales reps quote readership, and the owner's question is simpler than either: which of these produced jobs, and how much were they worth?

The usual answer is a shrug and a renewal. Print has a reputation as the channel nobody can measure, and the reputation suits everybody involved: the publication does not have to prove anything, and the business does not have to find out. Both are wrong about the facts. A print ad cannot be clicked, but it can be called, visited and named, and every one of those leaves a record that can be matched to a sale.

What you buy when you buy print space

A print ad buys space in an audience somebody else owns. You get a page, a position and a circulation figure, which is a count of copies distributed. You do not get the list of who received them, which is the single biggest difference between print advertising and mail you send yourself.

That difference decides the whole measurement method. With direct mail, the recipient file is yours, so you can match it against your sales and set aside a random holdout. With a newspaper, a magazine or a trade journal, the subscriber list belongs to the publisher, and no publisher hands it to an advertiser. There is no matchback in print. Measurement has to come from the response, not from the audience.

So the first thing to understand about a circulation figure is what kind of copies it counts, and the second is that the figure has been shrinking for twenty years.

A paid subscriber who chose the magazine, a copy mailed free to somebody who fits the publisher's profile, and a copy on a dentist's waiting-room table are three different exposures, and the headline number adds them up as one. Before you buy, ask for the audited statement, not the media kit, and look at the mix. A publication that is mostly paid copies in your service area is a different purchase from one that is mostly free copies spread across a region you do not serve.

By those figures, weekday circulation in 2022 was 20.9 ÷ 53.3 = 39% of its 2005 level, and advertising revenue was 9.8 ÷ 49.4 = about a fifth. Our reading of that is not that newspapers are a bad buy. It is that the generic case for print has collapsed, and what is left is a specific case: this publication, these readers, this service area. A specific case is exactly the kind that can be tested against sales.

Circulation tells you how many copies were printed and where they went. It does not tell you who read your ad, and it never tells you who bought. Use it to shortlist publications, never to judge them.

Readership figures are estimates on top of counts

Publishers and their trade bodies sell print on readership: the number of people who read a copy, not the number of copies. Readership is an estimate built on circulation, it is always larger, and it is a reasonable way to compare publications. It is not evidence that any reader became a customer.

Those figures describe the medium as a whole, put forward by the industry that sells it. They are not wrong to quote, and a magazine on a waiting-room table plainly is read by more than one person. But "take an action" covers looking something up, keeping the page, or mentioning the ad to somebody, and none of that is revenue. Your own ad in your own publication earns whatever your sales file says it earned. That is the only number you can budget on.

Give every publication a door no other channel uses

To measure a print ad, make the response to it identifiable: a phone number printed in that publication and nowhere else, a short web address leading to a page only print readers see, and a named option for the publication in your intake question. Each lead that arrives through one of those doors carries the publication as its source.

Three doors, because readers use all three, and each one catches people the others miss:

• A number per publication. A number printed only in the Sunday daily, a different one in the community weekly. Calls to it are print leads by definition, and the call log records the caller's number, which is the field that can be matched to a sale. If you can afford it, give a magazine its own number per issue; if not, per publication is enough. Our post on call tracking and revenue covers why the call log alone stops short of revenue.
• A print-only page. A short, typeable address — yourcompany.com/home, not a QR code alone — that lands on a page with a form and appears in no ad, email or menu. A form submitted there is a print lead with an email and usually a phone number.
• A named answer. "How did you hear about us?" with the publications listed by name, not one option labeled "Newspaper or magazine". This is the door that catches the reader who kept the page, came back weeks later and phoned your main line. Our post on measuring word of mouth explains how to word the question so it records sources rather than routes.

Whatever door the lead came through, write the publication into the same source column every other lead uses, with the same spelling every time. "Herald", "The Herald" and "Herald Sunday" will report as three channels unless somebody decides they are one. Decide before the first ad runs.

One trap deserves its own sentence: a tracking number that forwards to your main line is only useful if your team answers the phone and records the lead the same way regardless of which number rang. The number identifies the source; the lead record, with the caller's number on it, is what makes the source countable against sales.

The arithmetic: cost per matched sale, by publication

Once print leads carry their publication, match them to closed sales and divide. Revenue from matched sales over spend gives each publication's return, and spend over matched sales gives its cost per sale. Those two numbers routinely rank publications differently from cost per thousand copies, which is the number print is sold on.

Here is the remodeler's quarter, with every lead exported from all three doors and matched against closed jobs.

The arithmetic: the daily delivered 85,000 Sunday copies six times, or 510,000, so its cost per thousand is $14,400 ÷ 510 = $28.24. The weekly delivered 12,000 copies thirteen times, 156,000, at $3,900 ÷ 156 = $25.00. The magazine delivered 30,000 copies three times, 90,000, at $10,500 ÷ 90 = $116.67. On matched sales, the daily returned $63,000 ÷ $14,400 = 4.4, the weekly $42,700 ÷ $3,900 = 10.9, and the magazine $19,000 ÷ $10,500 = 1.8.

On cost per thousand, the daily and the weekly look like the same buy and the magazine looks four times overpriced. On matched revenue, the weekly earns two and a half times what the daily earns per dollar, because its 41 leads closed at 7 ÷ 41 = 17%, against the daily's 9 ÷ 96 = 9.4%. A smaller paper read by neighbors of the people who buy is often worth more than a bigger one read across a metro area you only partly serve. That is an argument, and the table is how you test it for your own business.

The magazine is the hard case, and the place to be careful. Two sales is not a result; it is two sales. A full-page ad in a magazine that sits on coffee tables for months, sold to people planning a project that takes a season to decide, will keep producing leads after the quarter closes. Before judging it, export leads covering the whole period the ad could have produced them — our post on attribution windows sets out how to choose that period — and wait at least one full sales cycle after the last issue.

Cost per thousand copies is how print is sold. Cost per matched sale is how it should be bought, and the two often disagree about which publication is the bargain.

The door catches a floor. The rest shows up elsewhere

Many readers who act on a print ad never use the printed number or address. They search your name, call the number they already have, or walk in. Their sales land in branded search, in other channels or in Direct / Unknown, so matched print revenue is a floor and the rest has to be estimated by comparison.

This is the same problem every no-click channel has, and we went through it for billboards: the dedicated number captures the people who acted on the ad directly, and a search for your name collects the credit for the others. In print it is, if anything, larger, because a page can be kept. Somebody tears out the ad in March and searches for you in June, and by June nobody remembers the paper.

Without a subscriber list there is no person-level holdout, so the comparison has to be by place or by time:

• By zone or edition. Many dailies sell zoned editions and many inserts are distributed by ZIP code. Run the ad in some zones and not in comparable ones, and compare branded search, calls to your main line and total closed revenue in each, before and during the run.
• By schedule. Run the ad for eight weeks, pause for eight, run again. If closed revenue in the circulation area moves with the schedule, beyond what the dedicated doors captured, that is the rest of the effect. If it does not, the floor is close to the whole of it.
• By watching where the credit goes. If branded search leads and Direct / Unknown sales in the circulation area rise when the ad runs, print is likely the reason, and those channels' figures in those weeks are borrowed.

Design the comparison before the first insertion, not after a good month. Our post on incrementality tests covers how to choose the areas and how large a difference has to be before it means anything. And whatever you find, do not take the unmatched sales and spread them across your channels to make print look complete; we explain why that inflates the wrong lines in unattributed revenue.

Inserts and coupons: closer to mail, measured like print

An insert rides inside the newspaper, distributed by zone or ZIP code, so it reaches households without giving you their names. A coupon on it measures the people who redeemed the coupon, which is a small and unrepresentative share of the people it moved.

Inserts sit between mail and print. You choose the areas, as you would for a saturation mailing, but the publisher delivers to its own subscribers and you never see the list. That makes them a natural fit for the zone comparison above: an insert in some ZIP codes and not in others is a geographic test you were going to run anyway.

Divide one by the other and about 0.81 ÷ 34 = 2.4% of coupons distributed in 2025 were redeemed, against 0.87 ÷ 53.1 = 1.6% in 2024. Those are grocery and consumer-goods figures across every format, not a local service business's insert. But the lesson travels: a coupon is redeemed by the reader who kept it and remembered to use it, and if the redemption count is your measure of the insert, nearly everybody else it reached does not exist. Use the code as one more door, write it on the lead with the customer's phone or email so it can be matched, and report redemptions as a breakdown of how people responded.

Trade journals: the reader is not the buyer

A trade journal ad is often read by an engineer, a specifier or an office manager who will never sign the order. The lead and the sale can carry different people, different phone numbers and different email addresses, months apart, so matching by person will miss sales the ad genuinely started.

For industrial manufacturers and other business-to-business sellers, the job of a trade journal ad is usually to put the company on a shortlist, not to produce a call that becomes an order. The engineer requests a datasheet; purchasing places the order eight months later from a different address. Person-level matching will connect the two only when the same phone or email appears on both, and it often does not.

Two adjustments make trade print measurable without pretending. Ask where the buyer first heard of you at the request for quote, not only at first contact, and record the answer as a source on that lead, with the buyer's own contact details. And export leads over the full buying cycle, which our post on long sales cycles shows is commonly longer than the reporting period anybody wants to wait for. A trade journal judged on a quarter is judged on the wrong quarter.

CloseRev matches people, not pages. If your lead file has a row for each print-driven lead — a call to a number printed only in that publication, a form on a print-only landing page, or a "How did you hear about us?" answer naming it — with the lead's phone or email and the publication in the source column, those leads are matched to closed sales like any other source and each publication reports under exactly the name written there. A sale with no such row stays in Direct / Unknown. It does not read coupon codes or estimate lift; the zone and schedule comparisons are yours to run.

What print cannot be measured for

Some of what print does leaves no record anywhere: the credibility of appearing in the local paper, the name a reader recognizes when a neighbor mentions you, the page kept for two years. No method recovers that, and an honest report says so rather than inventing a number for it.

It is tempting, when print's floor looks thin, to argue it up with brand value, and just as tempting, when the budget is tight, to cut everything without a matched sale. Both are claims without evidence. What can be said is narrower and more useful: this much revenue was matched to this publication; this much moved with the schedule or the zone; the rest is unknown. A law firm whose name has been in the same weekly for twenty years may well get cases from that presence that no door will ever catch. The right response is to say that in words, next to the measured floor, not to put a number on it.

Print does things nobody can measure. That is a reason to report them as unknown, not a reason to stop measuring the part that can be measured.

What to report for print

A print report has four lines per publication: spend and copies, leads by door, matched sales and revenue with cost per sale, and the result of any zone or schedule comparison. Readership and impressions belong in the media plan, not in the verdict.

1. Spend and copies delivered, from the invoice and the audited statement, with the paid share noted.
2. Leads by door: calls to the printed number, forms on the print-only page, intake answers naming the publication.
3. Matched sales, matched revenue and cost per matched sale, labeled as a floor, over a lead period long enough for the sales cycle.
4. The comparison, if one was run: revenue in the zones or weeks with the ad against those without, with the range and the honest caveats.

Run that for two quarters and the renewal conversation changes. The rep brings circulation and readership; you bring cost per matched sale by publication. Some publications will look better than you feared — community weeklies often do — and some will be on the list to pause for a schedule test. Either way, you will be deciding on your own sales, which is the only evidence about your print that anybody actually has.

Print is not the unmeasurable channel. It is the channel that was measured by the wrong number. Replace circulation with matched sales per publication, and the decision about what to renew makes itself.