Even a few years ago, Joseph Addison and Sir Richard Steele, those 18th-century London gallants and the founders of the Spectator, would have recognized the forms and modes of business that characterized our newspapers and magazines. Not now.

For 300 years, two related sources of revenues sustained publications: subscriptions and advertising. The system worked imperfectly. Most readers of newspapers and magazines were freeloaders, borrowing copies someone else had bought; and because no one really knew how many people read publications, or how advertisements influenced readers' purchasing, advertisers spent their monies inefficiently.

But so long as subscription and advertising revenues grew, the system did work. In turn, the business of publishing supported the profession of journalism, which was, when all is said and done, a useful thing. In open societies, magazines and newspapers were the most important exchanges in the free marketplace of ideas. Publications informed, instructed, diverted, and delighted.

But the Internet taught readers they might read stories whenever they liked without charge, and it offered companies more-efficient ways to advertise. Both parties spent less. As a consequence, today the business of media is sickly.

In recent months, the news about magazines and newspapers, distressing for many years, has become alarming. During the first quarter of 2009, the advertising revenues of newspapers declined, on average, 30 percent; in the last six months of 2008 (the most recent period for which we have reliable numbers), subscriptions fell by 7 percent. The number of ad pages in consumer magazines shrunk by 26 percent in the first quarter of the year; and while magazine circulations are not declining as rapidly as those of newspapers, it is becoming more and more expensive to maintain their rate bases (the circulation numbers from which publishers derive advertising rates), and with fewer advertisers willing to pay to reach those readers, a less and less rational investment.

Everywhere, newspapers and magazines are going broke. Sun-Times Media, the owner of 58 newspapers including the Chicago Sun-Times, declared bankruptcy at the end of March. The Star Tribune Holdings Corporation, the Journal Register Company, and Philadelphia Newspapers LLC are all, similarly, bankrupt. The Seattle Post-Intelligencer now exists only on the Web. The Rocky Mountain News, Colorado's oldest newspaper, is gone. The business magazine Portfolio, upon whose launch Condé Nast lavished more than $100 million, is gone. PC Magazine, gone. Domino, gone. Country Home, gone. It's a dolorous toll.

What can be done to save our surviving newspapers and magazines? Among those who write about new media, a fashionable wisdom has emerged, expressed most energetically by Clay Shirky, a professor at New York University. In "Newspapers and Thinking the Unthinkable," a much-distributed post on his blog, he writes, "Round and round [it] goes, with the people committed to saving newspapers demanding to know 'If the old model is broken, what will work in its place?' To which the answer is: Nothing. Nothing will work."

The Götterdämmerung-of-mainstream-media argument has a weak and a strong formulation. Shirky himself is an eloquent exponent of the gentler version. He argues, "Society doesn't need newspapers. What we need is journalism." Shirky believes that the coming decades will see a variety of nonprofit experiments whose funding sources will be similar to those that have sustained him as an academic, such as endowments, sponsorships, and grants. One day, some innovator will stumble upon something that will reliably subsidize the publications of the future.

The strong version is most associated with Dave Winer, a grumpy California software programmer best known for helping to develop the Web-feed format RSS and for his blog, Scripting News. Winer has written, and not without glee, "Fifteen years ago I was unhappy with the way journalism was practiced in the tech industry, so I took matters into my own hands. And then dozens of people did, and then hundreds followed, and now we get much better information about tech. It will happen everywhere, in politics, education, the military, health, science, you name it. The sources will fill in where we used to need journalists. ... Everyone is now a journalist."

If media companies can't earn money, and everyone is a journalist, it follows that "amateurs" (Shirky) and "sources" (Winer) will be part of a "decentralized" media (Winer), whose stories will be distributed by "excitable 14-year-olds" (Shirky).

This is all folly and ignorance. Shirky, Winer, and other evangelists know nothing about the business of media. True, the journalists who write about these matters for mainstream media often know as little; I didn't understand much until I became the publisher of Technology Review as well as its editor in chief. But Shirky and Winer are disgruntled consumers and, as bloggers, advocates for an insurrection. Thus, they are to be read skeptically. Their prescriptions would be more convincing if they were less polemical and better informed by some knowledge of what publishers sell.

Shirky and Winer share the conviction that media-as-a-­business, with its attendant professional writers, editors, art directors, directors of consumer marketing, and advertising salespeople, is being overthrown by ordinary people, using digital technologies. That's because they conflate mainstream media with printing presses. As Shirky explains, "Printing presses are terrifically expensive to set up and to run. ... [But] the competition-deflecting effects of printing cost got destroyed by the internet, where everyone pays for the infrastructure, and then everyone gets to use it."

For decades, most print publications have cheaply rented presses owned by third parties--but let that go. The printing press stands here as an objective correlative for the material production and distribution of media. Shirky and Winer's real error is that the physical is the least of it. The comparative advantage of mainstream media is not the ownership of presses, but the collaboration of professionals. The creation of good journalism is a tremendously laborious process, requiring an infrastructure more expensive than any press. The illustration and design of stories has an infrastructure, too. Developing an audience that will attract particular advertisers requires another infrastructure. Selling advertising requires yet another. These structures, which allow publications to reach large, coherent audiences, can exist only within complex organizations, mostly businesses.

Some of those structures must be reinvented for the Internet. Others, particularly editorial, still work well. I am sure of this, because the number of people who read newspapers and magazines is growing. Of course, with few exceptions that growth is all digital. To take one example, between 14 million and 22 million read nytimes.com every month; the print circulation of the weekday Times is just one million. In all, on any day, 32 million Americans read their news online. Those numbers suggest contented customers. Of course there is a good business for mainstream media in electronic publishing. The absorbing question is how to pay for what pleases so many.

It is a canard that neither mainstream media's managers nor its journalists have good answers to that question. There are plenty of stupid publishers and editors, and their publications will die; but there are many smart, technology-savvy leaders, too, and their publications will prosper. While the details are still debated, the broad outlines of tomorrow's media are becoming clearer. Consumers must pay for more of what they read; publishers and the media buyers who purchase advertising must be given technologies that will make online display ads more competitive with the keyword ads that search firms sell. Some of the things that must be done cannot be done by the media itself; it won't be easy, and it might not happen, but it can be done.

Below is my prescription for saving magazines and newspapers. Publishing is an involved and complicated business, and the following points are, perhaps, tediously interrelated and technical. The details will be of interest mainly to media professionals, but since it is detail that has been lacking in similar prescriptions, specificity is valuable. (A more concise, generalizing version of this column was published in the May issue of Technology Review and can be read online here.)

How to Save Media

A. Circulation, subscriptions, platforms, and frequency

1. Print is dying but is not yet dead. Magazines and newspapers need a strategy for print even as they prepare for the digital platforms of the future. Publishers should allow rate bases to decline to organic levels, defined as circulations where 65 to 75 percent of readers renew every year with minimal consumer marketing. Paid circulations should be much smaller, and the foundation of new, electronic subscriptions (see A.3, A.4, and A.5).

2. For many decades, publications were overdistributed to readers who didn't really want them, because publishers were former ad salesmen who hoped to profit by charging advertisers the highest possible rates (see B1). Subscribers became used to paying too little, because advertising revenues underwrote circulation costs. Publishers should charge fewer readers (see A.1) more for subscriptions. Further, each subscription should be healthily profitable within two to three years, recouping the costs of customer acquisition.

3. Content that some readers pay for in one medium (now, usually print) should never be offered without charge to other readers in another medium (usually electronic). Instead, publishers should distribute editorial to their subscribers on a variety of platforms (see A.5 below). This is not to say that much content should not be freely available to readers and paid for by advertising revenues. (To learn what should be paid and what free, see C.3 below.)

4. Just as publishers shouldn't worry about what platforms their readers prefer so long as that platform is profitable, they should offer readers as much choice in subscriptions as is rational. A reader should be able to buy a lifetime's subscription or subscribe for a year, a month, a week, or a day. If it made sense, a reader should be able to buy a package of stories, or even one story. The price of a subscription should reflect its duration and the platforms on which it is delivered.

5. The most important publishing platform of the future will probably be lightweight, thin, flexible screens that use electronic ink. That's because the editorial distributed to such screens will be as interactive as that on today's websites yet retain the fonts, graphical design, and illustrations and photographs of traditional media (a wonderfully rich visual grammar that art directors labored over for centuries). But publishers must not become fixated on platforms; they must regard them as mere distribution channels favoring different kinds of content. Again, publishers should offer their readers as much choice as is reasonable. Over the next decade, they should distribute editorial content to personal computers over today's Web, to small devices like the iPhone, to larger devices like Amazon's Kindle, to electronic-ink devices as they emerge, and to print publications (at least for a little longer).

6. Printing and physical distribution are expensive. For as long as they still print and mail publications, publishers should do it less frequently. Monthly magazines can be printed bimonthly; weekly magazines can be printed biweekly; newspapers can print on weekends only.

B. Advertising, sponsorships, and classifieds

1. Since the founding of the Spectator, publishers have overcharged advertisers for space in print publications, both by overdistributing their publications (see A.2) and by delivering readers who were indifferent to advertisers' messages (see B.3). Thus, advertisers are behaving rationally when they buy online advertising, which is more efficient and more easily measured. Rather than resenting advertisers, or hoping that they will somehow recover their enthusiasm for inefficiency and unaccountability, publishers should offer their media partners better return on their investments. Neither resentment nor hope are business strategies.

2. Today, the online ads most loved by advertisers are keyword or search advertisements (the sponsored links that appear near search results on Google.com and search sites, or that are sold by search firms and appear on other sites). Advertisers love keywords because their effectiveness is unambiguous: advertisers pay directly for click-throughs or transactions. Spending on keywords grew 21 percent in 2008, mostly at the expense of print, local television and radio, and Yellow Pages advertising; it now constitutes 45 percent of all online advertising. By comparison, the display or banner advertising that media companies sell grew only 4 percent the same year. This is a problem, because display advertising was meant to fund the great shift of readers to new media. But banner advertising will compete with keyword ads only when there are better audience measurement tools. Amazingly, today no one knows how many people visit websites. No established third-party supplier of audience measurement data is trusted. Internal Web logs exaggerate audiences. Yet better audience measurement would allow advertisers and media buyers to do cross-media comparisons, and it would benefit the vast majority of smaller media sites that don't have large audiences. Trusted audience figures could then be supplemented by measurement of click-throughs or other kinds of engagement measurements, thus making display advertising more valuable to advertisers. Happily, companies including Google and the California startup Quantcast are working to measure Web audiences in new and clever ways. (See "But Who's Counting?" in the March/April 2009 issue of Technology Review.)

3. In addition to being clouded by disagreement about the size of Web audiences, banner advertising suffers from deep structural problems that must be addressed before advertisers will spend really large sums. These are various and daunting, but they all involve, in one way or another, the absence of commonly accepted, automated means to create, sell, serve, and track the performance of display ads. Again, a number of companies are working to solve these difficulties.

4. Among the most promising advertising forms for media companies is custom advertising. In these arrangements, a publisher works directly with an advertiser and its agency to create a unique campaign, attached to a particular editorial event, that targets a publisher's audience and integrates all the publisher's platforms, often with a microsite that harvests sales, leads, or whatever else the advertiser values. Technology Review has benefited from such custom advertising with the governments of Spain and Singapore, and with companies such as Microsoft and Pitney Bowes. But the problem with custom advertising is that many advertisers, emboldened by a depressed advertising market, wish to blur the lines between editorial and advertising. Clear guidelines from industry associations like the American Society of Magazine Editors (ASME), as well as a long tradition, established rules about how advertisers and publishers should work together in print; if publishers are to retain the trust of their readers, they must have similar rules for electronic media. In the long run, such rules will benefit advertisers, too, by preserving the audiences they wish to reach.

5. Classifieds, except in the very narrow sense of job listings in professional publications, are no longer part of the business of publishing. Get over it.

C. Editorial

1. Editors can charge readers for content that is uniquely intelligent; that relies on proprietary data, investigation, or analysis; that helps readers with their jobs, investments, or personal consumption; or that is very expensively designed. Everything else should be available free, because it is news or opinion, which are commodities and must be offered up to the aggregators, social networks, and feeds. Such content can be monetized (to use the ugly jargon of our industry) only through traffic, which drives ad impressions. Here, although the quality of the editorial should meet the minimal standards of a publication, editors shouldn't invest too much time or money: good enough is best.

2. Mostly, editors should give readers what they say they want. This will take a sea change in the attitudes of editors. Indeed, my own feelings in this matter have been overturned. As I rose through the editorial ranks of various magazines, I was encouraged to cultivate a mild contempt for readers. We disdained the market research our publishers commissioned, telling ourselves that readers didn't know what they wanted. But electronic media and social technologies have had a paradoxical effect: on the one hand, disappointed readers can abandon a publication with a click of a mouse or stab of a thumb, and at the same time they have strengthened readers' proprietorial sensibilities. Hence, our rule at Technology Review has become that about 80 percent of what we publish should satisfy our reader' expectations, and the rest can blow their minds.

3. Update 5.5.2009. One of the the things that some readers say they want is to be able to post comments about stories as well as their own stories to the Web sites of media companies. Often, such readers want to be able to communicate directly with one another, using social technologies. The readers who want to do this are not very many, but they feel strongly about the subject, and become angry if they suspect editors wish to be "gatekeepers". Editors must welcome such readerly participation, and should open their editorial departments to the wider world. (For myself, I love to know what our readers think and know.) Considered only as a business opportunity, reader-generated content make sense: the content itself is free, the associated costs are limited to Web hosting and developing the functionality whereby readers can post to sites, and the content drives traffic and ad impressions.

4. Editorial departments should become smaller. How small? Unless a newspaper or magazine has a deep-pocketed patron, it must turn a real and predictable profit. If it has a patron, a publication's losses must be predictable and sustainable. Along with other expenses, editorial budgets must retract until they are rational or the publication will be shuttered. Accepting this will be inordinately difficult for most editors; only their own termination or the bankruptcy of their company will really convince them.

Technology Review made many of these changes after I became publisher in 2004. The rest, we will be making during the next few months. The coming years will demand yet more changes that I cannot anticipate. It has all been very unpleasant, and the unpleasantness will only continue. But this is what is necessary. For publishers to wish that reality were otherwise is the height of fecklessness; it is, in fact, a species of madness.

Things change or die, including once-cherished organizations. Today's newspapers and magazines will be transformed or replaced by other publications, which will have new forms and modes of business. There will be a great and terrible clearing: scores of newspapers and magazines will vanish; those that survive will be much reduced; and most people employed as journalists or media professionals today will have different jobs in five years. At the same time, millions of Shirky's amateurs and Winer's sources will flourish to bewitch readers. But anyone who tells you that media-as-a-business is dying is wrong.