Follow the Numbers

The current travail of the Boston Globe got me wondering: The big regional metros that are the most threatened in this environment from the loss of classified advertising -- we talk about them as if we know exactly what is happening to their sales. For example, anyone discussing Boston will comment on its being a younger and more wired market than most, which is true, and say, thus, the Globe hurts.

But where exactly does the Globe sell and to whom? If the analysis is correct, the Globe should be falling through the floor in Suffolk and Middlesex Counties, which include Boston and Cambridge -- but, of course, those also include places such as Framingham and Revere. Perhaps it would be doing better in Norfolk and Essex, which include the more traditional suburbs of the North Shore and Brookline. Or perhaps not. Perhaps everyone in blue-collar South Boston still reads it and no one in upscale Newton does. I have no idea.

But I suspect almost no one writing pieces about whether the Globe should die does either. What they know is whether they and their friends read it, or whether they think anyone should read a print newspaper ever -- or whether everyone who now reads the Globe should simply read the Times, because, of course, the debate over online news always ends up being in part a debate about whether "news" is the same as "the A section of the New York Times."

Obviously people at the Globe know where their readers were and are. They can break it down by zip code.

What I can do is tell you where the Globe's circulation was in 1991, because among the things I have kept is a book with the 1991 circulation of every newspaper broken down by county.

First off, some comparison. A typical monopoly daily newspaper in 1991 was delivering to about 50 percent or more of the households in its home county. Columbus 51 percent, Cleveland 57, Peoria 54, Wichita 52, New Orleans 50, St. Louis 46, Portland 48. (In 1991 there were still a lot of morning-evening combos, so it's hard to find direct comparisons.) So by this time we were already far past the golden era when 70 or 80 percent of households took a daily newspaper. Newsday was hitting 55 percent of Long Island, so with the metros it may still have been true there. Of course, Newsday had this problem with reporting circulation.

In 1991, the Globe's daily circulation in its metro area broke down into two tiers. In Suffolk, Norfolk and Middlesex Counties -- the inner Boston area -- the Globe was listed as going to one out of three households. Obviously this includes home delivery and street sale, deliveries to offices and libraries, whatever, and particularly in a commuter city such as Boston includes papers sold downtown to people coming in from elsewhere, so the number is approximate. But it's probably largely accurate.

In the next ring of suburbs -- Essex and Plymouth Counties, or the North Shore and the far South Shore -- the Globe was doing 17 to 20 percent. It was doing the same level in Cape Cod and the islands, but you only had to sell 500 papers on Nantucket to get that level so this was not a big deal. The only other area where the Globe had any real strength was central and southern New Hampshire, where it got far suburbanites and people who did not want to read the Manchester Union Leader. I don't know Boston that well, but I expect that most of the urban-sprawl new housing that was being built was in Plymouth County and New Hampshire.

At this point, about 80 percent of the Globe's daily circulation was in metro Boston. On Sunday, each county figure went up by about two-thirds, except in Boston's county itself; in Norfolk, the Globe's best county on Sunday, the Sunday paper went to 51 percent of households. Because few people commute on Sunday, this probably is an accurate figure.

The other 20 percent of the daily circulation was in Cape Cod, the islands, Hampshire County -- UMass, Amherst, etc. -- and New Hampshire, plus marginal copies as far away as Maine and Connecticut. The Sunday Globe did better, of course, but mainly in New Hampshire and a number of readers in Vermont -- perhaps a lot of weekend homes. The Globe may have seen itself as the newspaper of New England, but it was essentially the newspaper of eastern Massachusetts and liberals in New Hampshire.

Within this area, of course, it was not only competing with the Herald, but with myriad small dailies, in Lynn and Waltham and Framingham and Quincy and Nashua and on and on, as well as weeklies. Essex County at that time had seven daily newspapers. The Globe couldn't sell local news as its main draw in such a market. It had to sell the A section, politics, features, the Sox -- and, of course, a lot of Help Wanted ads. The sort of stuff that everyone, not just newspapers, now offers free online, but that you weren't going to get much of from the Gloucester Daily Times.

Now, where has this circulation fallen off the most? In Suffolk, perhaps, where Boston is full of singles and DINKs? In Cambridge? (Middlesex County is incredibly diverse, from Harvard to old factory towns to 1950s suburbs.) New Hampshire? Has the falloff been even across the board? I have no idea, and thus no theories. But I think it would be useful to know. For example, if we were to find that household penetration in Suffolk and Norfolk was not that much different, it might mean one thing. If we were to find that sales in Suffolk were off by 50 percent but most of that was street sales in downtown Boston, it might mean something else. If we found that the collapse has been strongest or weakest in New Hampshire it might mean something else. If it simply happened across the board everywhere, something else again.

But I remember some conversation at my paper about insert advertisers not wanting to do business with you unless your county penetration was some figure -- 28 percent, 30, 33, I don't remember exactly but it was around there. By this standpoint, the daily Globe was already becoming marginal to such advertisers. In 1991 it deliverered 34 percent of its home county. The Globe probably was publishing a profitable weekly with a six-day come-on offer. That may also have been the case in Charlotte or Wichita, but the business model was not as tenuous. But as long as that Sunday Globe was the main pipeline for classified advertising to 51 percent of largely upscale Norfolk County, they could charge whatever ad rates they wanted and staff accordingly.

Just out of curiosity I looked up another troubled paper, the San Francisco Chronicle. At this point it was still in a JOA with the Examiner, but by that point the Examiner was not a factor outside of the city of San Francisco. Even with that, the Chron had 40 percent penetration in San Francisco County, which is the same as San Francisco City. The same sort of pattern as for the Globe followed -- penetration in the 30s in the closest suburban counties, Marin and San Mateo; in the high teens in Alameda, Contra Costa and Solano, which are Oakland, Berkeley, Vallejo and lots and lots of new suburbs. By the time one got to San Jose, or Napa, the Chron was down to 6 percent of the market. (That's right, the Chron in 1991 was read by almost no one in San Jose, which is, what, 40 miles away? The Merc hit 45 percent.) But on Sunday they had 51 percent of Marin County. Talkin' 'bout Mill Valley, that's my home....

But metro area circulation accounted for only 57 percent of the Chron's figures. The paper was huge throughout northern California, getting a significant number of households almost all the way to Oregon and as far away as Carson City, Nev. By "significant" I mean "about the same number, or more, as it was getting in San Jose." So in discussing the ills of the Chron, I would want to know -- does no one read it in print in the city of San Francisco anymore? Or do they? Is it still read in Marin and San Mateo, which include the older upscale suburbs such as Sausalito and Burlingame? Or has the falloff been in Butte and Mariposa and Shasta and Siskiyou counties? Or, on the other hand, do the residents of those more traditional areas still love their Chron (and they are the main audience it has left)? Or has the Chron simply stopped trucking papers all the way to Nevada?

It's not that figures provide all the answers. But in these convoluted and diverse metro markets, it would be nice to know where the problems really are -- as opposed to anecdotes from people who are too cool to read newspapers.

Why This Could Be Good for Newspapers

In the 1980s newspapers started increasingly hearing from advertisers that they wanted to know who newspapers' readers were. This flummoxed them. Newspapers' readers were everybody. Merthe's and Vaughn-Ragsdale and Gold's had never asked them who their readers were. (A gold star to anyone who posts all three cities of those department stores.)

Breaking down circulation by county, newspapers had been doing for years. By zip code was harder. For smaller newspapers, zip codes had never mattered. In big cities it mattered to advertisers a great deal, but delivery routes were not by zip code. No one thought in those terms. Some newspapers had no idea who their customers were. One of the reasons for the Detroit newspaper strike of 1995 was that the drivers and distributors controlled the circulation lists and the newspapers couldn't say who was getting the paper. None of this had mattered when the newspaper said "We have 600,000 circulation!" and Hudson's said, "Great! We'll buy 15 pages on Sunday!"

But in the 1970s, advances in printing technology made it possible to produce a specialty publication for any interest, just as changes in the number of storefronts available, a more spread-out population, and a greater inventory of goods on offer made it impossible for a department store to offer a representative sample of everything. So you had Cat Litter Box Scoop World, and advertisers of litter box scoops and anti-furball medicines knew who was reading it and targeted their advertising.

Newspapers said: Everyone reads us! Advertisers said: We don't want everyone! Tell us why they read you. Newspapers said: Because they want to read the news! Advertisers said: Why is someone "reading the news" interested in my product? Newspapers said: Because we have 600,000 circulation! And things just drifted along, with newspapers raising rates and depending more on classified revenue to make up for their losses to advertisers, who felt that huge portions of their ad budget were being wasted on readers who had no interest in their product and would have none, ever.

Last week I noted three studies -- by Gallup, Pew, and Lee -- that seemed to indicate that at least at the moment, there is a floor of print-only readership somewhere in the area of 30 percent at bottom and 40 percent at top, even among 18-to-29-year-olds; add in print-and-online and it gets better.

The disadvantage to newspapers is that they can no longer shout: We have 600,000 circulation!

The advantage is: In an era of unlimited choice, these are people who have chosen to take your publication. Why? What are they looking for? How can that audience be marketed? They're no longer buying the paper to play the numbers based on Cha Ching or check whether AT&T went up .02. Among those readers, your newspaper has a niche of some sort. What is it? Who are they?

The median age of TV viewers is now 50. For newspapers, it is above 50. The median age of the U.S. population is 38. However, that includes people below 18. Readership studies always are of people 18 and above. What is the median age of people 18 and above? (Tell me if you can find this figure, I can't; but there are indications it's about 48.) But 75 percent of the U.S. population is above 18. To what degree does the age level of readership of newspapers match the average age of people 18 and up? (In other words, if you drop people under 18, if the average age is then 48, if the average age of newspaper readership is, say, 51, it would stack up nicely.) And are newspapers really trying to sell themselves to people under 18? Have they ever been? As the story on the age of TV viewers notes, the average age of Nickelodeon viewers is 10. It's useful in the TV world to know those figures; does including under-18s make any sense for newspapers?

People under 18 are not newspapers' target readership, in print or online. People over 65 skew the readership data because of their heavy reliance on print -- caused probably as much by not needing to waste time at work as their dislike of computers. Newspapers need to sell their reach, whatever it is, among people 18 to 65 and figure out a way to avoid having their penetration in this market be skewed out of recognition by both the absence of young people and the dominance of senior citizens, whose massive use of newspapers makes it too easy for people to dismiss the medium as simply Lawrence Welk on paper.

Newspapers' online strategy has been flawed by trying to reproduce on the Internet what they are in print, while not recognizing that while in print it makes sense to have a little bit of everything, the online world in itself is a little bit of everything. (Department stores had the same problem once malls became enclosed; an enclosed mall is in essence a department store, so what is the point of a department store in a department store?) But even so, newspapers can use declining readership as an opportunity to either define the readers they have, or increasingly try to market themselves to specific readers they have and forget others. This is happening with the geographic pullbacks but it doubtless needs to go further. And yes, it means trying to figure out who your target market is, instead of simply saying: It's everybody. Nobody wants everybody.