[This story is so remarkable, that I'm posting it here. As soon as the New York Times sends me a cease and desist, I'll remove it. But the article is brilliant. And if you find this blog even remotely interesting, do not pass "Go" until you've read it. Peter. 12/5/04]

The New York Times
December 5, 2004
The Hidden (in Plain Sight) Persuaders
By ROB WALKER

Over the July 4 weekend last summer, at cookouts up and down the East Coast and into the Midwest, guests arrived with packages of Al Fresco chicken sausage for their hosts to throw on the grill. At a family gathering in Kingsley, Mich. At a small barbecue in Sag Harbor, N.Y. At a 60-guest picnic in Philadelphia.

We know that this happened, and we even know how various party guests reacted to their first exposure to Al Fresco, because the Great Sausage Fanout of 2004 did not happen by chance. The sausage-bearers were not official representatives of Al Fresco, showing up in uniforms to hand out samples. They were invited guests, friends or relatives of whoever organized the get-togethers, but they were also -- unknown to most all the other attendees -- ''agents,'' and they filed reports. ''People could not believe they weren't pork!'' one agent related. ''I told everyone that they were low in fat and so much better than pork sausages.'' Another wrote, ''I handed out discount coupons to several people and made sure they knew which grocery stores carried them.'' Another noted that ''my dad will most likely buy the garlic'' flavor, before closing, ''I'll keep you posted.''

These reports went back to the company that Al Fresco's owner, Kayem Foods, had hired to execute a ''word of mouth'' marketing campaign. And while the Fourth of July weekend was busy, it was only a couple of days in an effort that went on for three months and involved not just a handful of agents but 2,000 of them. The agents were sent coupons for free sausage and a set of instructions for the best ways to talk the stuff up, but they did not confine themselves to those ideas, or to obvious events like barbecues. Consider a few scenes from the life of just one agent, named Gabriella.

At one grocery store, Gabriella asked a manager why there was no Al Fresco sausage available. At a second store, she dropped a card touting the product into the suggestion box. At a third, she talked a stranger into buying a package. She suggested that the organizers of a neighborhood picnic serve Al Fresco. She took some to a friend's house for dinner and (she reported back) ''explained to her how the sausage comes in six delicious flavors.'' Talking to another friend whom she had already converted into an Al Fresco customer, she noted that the product is ''not just for barbecues'' and would be good at breakfast too. She even wrote to a local priest known for his interest in Italian food, suggesting a recipe for Tuscan white-bean soup that included Al Fresco sausage. The priest wrote back to say he'd give it a try. Gabriella asked me not to use her last name. The Al Fresco campaign is over -- having notably boosted sales, by 100 percent in some stores -- but she is still spreading word of mouth about a variety of other products, and revealing her identity, she said, would undermine her effectiveness as an agent.

The sausage campaign was organized by a small, three-year-old company in Boston called BzzAgent, but that firm is hardly the only entity to have concluded that the most powerful forum for consumer seduction is not TV ads or billboards but rather the conversations we have in our everyday lives. The thinking is that in a media universe that keeps fracturing into ever-finer segments, consumers are harder and harder to reach; some can use TiVo to block out ads or the TV's remote control to click away from them, and the rest are simply too saturated with brand messages to absorb another pitch. So corporations frustrated at the apparent limits of ''traditional'' marketing are increasingly open to word-of-mouth marketing. One result is a growing number of marketers organizing veritable armies of hired ''trendsetters'' or ''influencers'' or ''street teams'' to execute ''seeding programs,'' ''viral marketing,'' ''guerrilla marketing.'' What were once fringe tactics are now increasingly mainstream; there is even a Word of Mouth Marketing Association.

Marketers bicker among themselves about how these approaches differ, but to those of us on the receiving end, the distinctions might seem a little academic. They are all attempts, in one way or another, to break the fourth wall that used to separate the theater of commerce, persuasion and salesmanship from our actual day-to-day life. To take what may be the most infamous example, Sony Ericsson in 2002 hired 60 actors in 10 cities to accost strangers and ask them: Would you mind taking my picture? Those who obliged were handed, of course, a Sony Ericsson camera-phone to take the shot, at which point the actor would remark on what a cool gadget it was. And thus an act of civility was converted into a branding event.

This idea -- the commercialization of chitchat -- resembles a scenario from a paranoid science-fiction novel about a future in which corporations have become so powerful that they can bribe whole armies of flunkies to infiltrate the family barbecue. That level of corporate influence sounds sure to spark outrage -- another episode in the long history of mainstream distrust of commercial coercion and marketing trickery. Fear of unchecked corporate reach is what made people believe in the power of subliminal advertising and turn Vance Packard's book ''The Hidden Persuaders'' into a best seller in the 1950's; it is what gave birth to the consumer-rights movement of the 1970's; and it is what alarms people about neuroscientists supposedly locating the ''buy button'' in our brains today. Quite naturally, many of us are wary of being manipulated by a big, scary, Orwellian ''them.''

In this case, however, it is not just ''them.'' It turns out that Gabriella and the rest of the sausage agents are not paid flunkies trying to manipulate Main Street Americans; they are Main Street Americans. Unlike the Sony Ericsson shills, Gabriella is not an actress. She is an accountant, with full-time work and a 12-year-old daughter, living in Bayonne, N.J. Aside from free samples, she gets no remuneration. She and her many fellow agents have essentially volunteered to create ''buzz'' about Al Fresco sausage and dozens of other products, from books to shoes to beer to perfume. BzzAgent currently has more than 60,000 volunteer agents in its network. Tremor, a word-of-mouth operation that is a division of Procter & Gamble (maker of Crest, Tide and Pampers) has an astonishing 240,000 volunteer teenagers spreading the word about everything from toothbrushes to TV shows. A spinoff, Tremor Moms, is in the works. Other marketers, particularly youth-oriented firms, have put up Web sites recruiting teenagers to serve as ''secret agents.''

Given that we are a nation of busy, overworked people who in poll after poll claim to be sick of advertisers jumping out at us from all directions, the number of people willing to help market products they had previously never heard of, for no money at all, is puzzling to say the least. BzzAgent, which has a particularly intense relationship with its fast-growing legions of volunteers, offers a rare and revealing case study of what happens when word-of-mouth theory meets consumer psychology in the real world. In finding thousands of takers, perfectly willing to use their own creativity and contacts to spread the good news about, for instance, Al Fresco sausage, it has turned commercial influence into an open-source project. It could be thought of as not just a marketing experiment but also a social experiment. The existence of tens of thousands of volunteer marketing ''agents'' raises a surprising possibility -- that we have already met the new hidden persuaders, and they are us.

Dave Balter, the 33-year-old founder of BzzAgent, is a smart guy, but he would be poorly cast as a slick, Madison Avenue mastermind. He's fresh-faced, good-humored, almost goofy. And he will cheerfully tell you that he has no definitive explanation for the number of average citizens who want to be, in company parlance, BzzAgents. In the beginning, he had a theory about what would motivate average citizens to generate word of mouth for his clients, but that theory was full of holes. It assumed, for instance, that agents would require some kind of quasi-financial motivation to do legwork for consumer companies.

Dave Balter's background was in loyalty marketing -- those frequent-flier-style programs that give rewards to dedicated users of a particular product, service or credit card. He read up on word-of-mouth-marketing theory, raised some money, hired a right-hand man and put the word out among family and friends that he was looking for ''agents.'' The idea was to build a network of people who would get points for spreading ''honest word of mouth'' and could cash in the points for cool products. ''The whole concept,'' he said, ''was rewards, rewards, rewards.''

The first full-fledged Bzz campaign was for a book called ''The Frog King.'' It lasted one month and focused on New York City. Balter persuaded Penguin Publishing to let him do it by charging the publisher nothing.

''The Frog King'' was a quirky, comic first novel by a young writer named Adam Davies. Davies had some connections in New York publishing (including Liz Smith, the gossip columnist), but he wasn't exactly going to get a giant publicity blitz. ''We didn't expect much'' from the buzz campaign, recalled Rick Pascocello, a Penguin vice president.

The guide for the agents, a no-frills seven-page document in those early days, welcomed them as members of ''an elite group'' of word-of-mouth spreaders and listed the contact information for ''your BzzLeader,'' BzzAgent JonO. (That was Jon O'Toole, Balter's right-hand man.) It summarized some of the novel's highlights, noting a few passages in particular that might be useful ''conversation points,'' and suggested tactics like reading the book on mass transit with the cover clearly visible, posting a review on Amazon.com and calling up bookstores and chatting with the clerk about this great new book about New York publishing with lots of sex and drinking whose title you can't quite recall. JonO signed the cover letter assuring agents that the folks back at the hive found the book laugh-out-loud funny.

Local events for ''The Frog King'' drew larger-than-expected crowds of 100 or 150 people, according to Pascocello, who said that thanks to the word-of-mouth campaign, the book sold in three months what he had hoped it would sell in a year. There are now more than 50,000 copies of ''The Frog King'' in print, and it's still selling. BzzAgent has had a steady flow of paying clients ever since (including Penguin, which has used BzzAgent to promote other books, like the novel ''The Quality of Life Report''). The fee it charges varies according to the size and nature of the campaign, but Balter said a 12-week campaign involving 1,000 agents would now cost $95,000.

BzzAgent has fewer than two dozen paid employees, though it is growing and recently moved to a larger office. These people are mostly young, without backgrounds in traditional marketing. When the company takes on a new client, they huddle to figure out whatever is most buzzable about the product at hand. This summer, for instance, they handed around and discussed a new line of Johnston & Murphy dress shoes, which feature a fiberglass shank, rather than a traditional metal one, so they won't set off metal detectors at airports. A whiteboard was filled with suggested conversation starters and likely sites for word-of-mouth opportunities, which later was transferred to a slick ''Bzz Guide'' for the agents.

As the number of agents has grown, the company can meet increasingly specific requests for, say, agents of a certain age or income level, or who live in certain parts of the country. It has done campaigns for a wide array of goods, and for major companies and brands like Anheuser-Busch, Lee Jeans, Ralph Lauren, even DuPont. Recently the company has also begun working with clients to begin converting existing loyal customers into private, well-organized, word-of-mouth missionaries.

Although Balter says he was pleased with his agents' efforts from the start, he did worry early on that the system could not be sustained. The problem was that while agents were spreading buzz and thus earning and piling up points, most were not cashing them in. That is, they weren't bothering to collect their rewards. ''We've built a broken model,'' Balter remembers thinking. He asked his colleagues from his loyalty-marketing days: Is it that the rewards aren't good enough? Are they too hard to get? After many hours of listening to the conflicting analyses of experts, he and O'Toole decided to ask the agents themselves about the points. ''We didn't realize the agents would want to talk to us,'' Balter said. This was another miscalculation; many of the agents very much did want to talk. In essence, they told Balter that there was nothing wrong with the rewards; it was just that the rewards weren't really the point. Even now, only about a quarter of the agents collect rewards, and hardly any take all they have earned.

Karen Bollaert, who is 32 and lives in the Bay Ridge section of Brooklyn, was among the firm's earliest agents, and became one of its most effective. When she signed up for her first BzzAgent campaign -- ''The Frog King,'' in fact -- she was working with a pharmaceutical researcher, mostly doing paperwork, and thinking about finding a more fulfilling way to spend her days. Like everyone who signs up at the BzzAgent site, she was accepted.

During active stretches, Bollaert puts in between 5 and 10 hours a week talking up products and writing reports about her activities. (She has signed up for many campaigns, including a perfume called Ralph Lauren Blue, a line of jeans for Lee and something called No Puffery, a gel to soothe skin below the eyes.) What, I asked her, if not the potential to get some free prizes for effort, made her bother to volunteer with BzzAgent? First, she told me, she gets the chance to sample new products shortly before they hit the stores, so she gets to feel a bit like an insider. Second, she has always liked to give people her opinion about what she's reading or what products she's using, and BzzAgent gives her more to talk about. Third, if she does like something, then telling other people is helpful to them. So participating is both a chance to weigh in and be heard, and also something close to an act of altruism.

What Balter said he learned from his agents is that lots of people like to tell others what they are reading and what restaurant they've discovered and what gizmo they just bought. In his view, BzzAgent is simply harnessing, channeling and organizing that consumer enthusiasm. This is presumably why it's so easy, so natural, for someone like Karen Bollaert to work word-of-mouth efforts into daily life. When, for example, a friend mentioned to Bollaert that she would have to get up early after a late night out on the town, she brought up No Puffery. When a pharmaceutical representative visiting her office worried about looking lousy at a meeting she had to fly to, Bollaert mentioned No Puffery. At her grandfather's wake, ''a relative told me how well I was looking,'' she wrote in one report back to the BzzAgent hive, ''and I mentioned that No Puffery helped to keep me looking calm instead of puffy-eyed and as horrible as I felt.''

The endless chatter of American consumer life that BzzAgent has infiltrated is not simply a formless cacophony; it has its structures and hierarchies, which have been studied exhaustively for decades. Tremor, the Procter & Gamble word-of-mouth unit, which also does work for a variety of non-P.&G. clients, was founded four years ago with those structures in mind. A key Tremor premise is that the most effective way for a message to travel is through networks of real people communicating directly with one another. ''We set out to see if we could do that in some systematic way,'' Steve Knox, Tremor's C.E.O., said recently. He added a second, closely related premise: ''There is a group of people who are responsible for all word of mouth in the marketplace.'' In other words, some friends are more influential than others, and those are the ones who are chosen to join Tremor.

Who are they? Check out the word-of-mouth industry's favorite graph. The graph is meant to show the pattern by which ideas or products or behaviors are adopted, and it looks like a hill: on the left are the early adopters; then the trend-spreaders; the mainstream population is the big bulge in the middle; then come the laggards, represented by the right-hand slope. This is not new stuff -- Knox himself cites research from the 1930's, as well as the 1962 academic book ''Diffusion of Innovation,'' by Everett Rogers -- but it has become extremely popular over the past five years or so. Seth Godin, who wrote ''Permission Marketing,'' ''Unleashing the Ideavirus'' and other popular marketing books (and whose ideas partly inspired BzzAgent), uses it, as do dozens of other marketing experts. Malcolm Gladwell's ''Tipping Point'' made an argument about these ideas that was simultaneously more textured and easier to digest than most of what had come before (or since), and it became a best seller. But whatever the intentions and caveats of the various approaches to the subject, the most typical response to the graph is to zero in on the segment that forms the bridge over which certain ideas or products travel into the mainstream -- influentials, trend-translators, connectors, alphas, hubs, sneezers, bees, etc. Let's just call them Magic People.

Knox said that Tremor's approach to finding the Magic People is intensively researched. The company tries to isolate the psychological characteristics of the subset of influential teenagers, and has developed a screening process to identify them. The details of this are a secret, but as an example, Knox noted that most teenagers have 25 or 30 names on their instant-messaging ''buddy list,'' whereas a Tremor member might have 150. Tremor recruits volunteers mostly through online advertisements and accepts only 10 or 15 percent of those who apply. The important thing, Knox said, is they are the right kind of kids -- the connected, influential trend-spreading kind. Knox mentioned a focus group of Tremor kids in Los Angeles, where several teenagers showed up with business cards. Magic.

Janet Onyenucheya was chosen by Tremor, and she is pretty much what you picture when you picture a trend-influencer. She is 18, African-American, beautiful, smart and, on the day I met her, was wearing a really cool pair of sneakers. An intern at an independent music publishing company in Manhattan, she is preparing to enter the Berklee College of Music in Boston in the spring. She got involved with Tremor a couple of years ago, while attending LaGuardia High School.

Onyenucheya gets free stuff from Tremor, and sometimes even a small check for taking surveys and participating in focus groups. She got to vote on the design for a T-shirt for the 10th anniversary of the Vans Warped Tour and for the design of a Crest toothbrush. This past July, she was invited to an advance viewing of two television shows, ''Lost'' and ''Complete Savages,'' at the Millennium screening room in downtown Manhattan. There were about 70 teenagers there, and pizza and sodas for everybody. Onyenucheya particularly loved ''Lost.'' ''When I came home,'' she said, ''I immediately told my five closest friends, like: 'Oh, my God, you just missed the greatest shows. I got to go down to the Millennium and saw a show called 'Lost' and it was so good, and we have to watch it when it comes out.' And I felt like I had the upper hand. Like, 'You don't know what I know.' ''

By and large, the word-of-mouth literature tends to describe our influence and degree of connectedness as something hard-wired. Magic People like Onyenucheya are born, not made, is the idea, which is why companies spend so much effort developing psychological profiles to find them. But the BzzAgent experiment largely discards that premise. Its agents are not screened. They are not chosen. They simply sign up. They are all kinds of people, all over the country: a 50-something bookstore owner in suburban Chicago, a young housewife near Mobile, Ala., a college student in Kansas. Many are teenagers, or even younger. At least one is 86 years old. And yet, it seems, they are able to persuade.

Jason Desjardins is a regular guy, a good guy, accommodating and polite. Twenty-eight, slim, clean-shaven, with close-cropped hair, he is the dairy manager at a supermarket in rural New Hampshire, part of the same supermarket chain he has worked for since high school. While he was wearing a Brooklyn T-shirt when we met, the truth is he bought it at the Old Navy in the Concord mall, and has never been on an airplane or even traveled outside of New England. Jason Desjardins is sweet and guileless, but he is not, by any expert definition, a Magic Person.

Desjardins stumbled across a reference to BzzAgent online, and he was interested. How could this thing work? He signed up, and soon after, they sent him ''Purple Cow: Transform Your Business by Being Remarkable'' -- Seth Godin's most recent book at the time, which was written with a BzzAgent marketing plan in mind -- and his life changed. It's hard to overstate how enthusiastic Desjardins is about BzzAgent. He joined campaigns for several other books, as well as for a beer called Bare Knuckle Stout, a spam-blocking service called Mail-Block and, yes, Al Fresco sausage. He figures he spends about 10 hours a week either buzzing or writing reports about buzzing. I visited him at his apartment in Bradford, N.H. We were joined by his wife, Melissa, a pretty woman with a stylish haircut and a big smile, and their 2-year-old daughter. I wondered how Melissa felt about her husband spending so much time on a no-money hobby. In fact, she was thrilled. She said she thought it had made him more open to other people. He used to be the kind of guy who just hated to call a mechanic about a noise the car was making; he would wait until the car actually broke down and he had no choice but to bother someone about it. He was in a shell. But that has changed -- partly because of Melissa, Jason wisely interjected -- but also partly because of his involvement in BzzAgent.

For starters, Desjardins said, BzzAgent ''turned me on to reading.'' And having enjoyed ''Purple Cow,'' he wanted to do his best to spread the word. The Bzz guide suggested he call a bookstore. For a while, he put it off. He would look at the phone and tell himself, I can do this, and he would try to rehearse what he would say, and this would go on for 15 or 20 minutes. ''I thought: What have I got to lose?'' he said. ''I'm never going to see this person.'' And finally he called and pretended he did not know the name of Seth Godin's new book. ''He'll call anybody now,'' Melissa said, smiling.

He printed slogans from ''Purple Cow'' (''Be Remarkable or Be Invisible'') onto card stock and hung them where his fellow employees could see them. He posted reviews on Amazon. He started conversations with co-workers, customers, strangers. He submitted a rave review for a fantasy novel he was buzzing called ''Across the Nightingale Floor'' to The Concord Monitor, and it was published; there's a laminated copy of the review on the fridge. He wrote to the governor touting Mail-Block. At the grocery store, when a co-worker moaned about not liking her job, Desjardins practically turned into a motivational speaker, waving his hands and quoting from another book called ''Five Patterns of Extraordinary Careers,'' telling her that if she wasn't happy she needed to take control of the situation. ''She did end up finding another job after that,'' he observed. Desjardins is ranked the 45th most effective BzzAgent, out of 60,000 nationwide, and proud of it. He has learned to influence.

This was all good for Desjardins, but it complicates what we thought we knew about word-of-mouth influence. The whole premise of the Magic People is that the rest of us take our cues from them because they have some special credibility, in the form of reputation or expertise or connections. In April 2003, that premise was put to the test when BzzAgent began a 13-week campaign for a restaurant chain called Rock Bottom Restaurant and Brewery, which has about 30 locations around the country. This particular campaign was studied by two academics: David Godes, an assistant professor at Harvard Business School, and Dina Mayzlin, an assistant professor at Yale's School of Management. The experiment involved more than 1,000 subjects; some were devoted Rock Bottom customers, and the rest were BzzAgents -- none of them Rock Bottom loyalists, and only a few had even heard of the chain. Rock Bottom wasn't running any other significant marketing program at the time.

Sales increased markedly. Godes and Mayzlin found that, consistent with past research, word of mouth traveled more effectively when it was spread not through close friends but through acquaintances (meaning that networkers -- the people with the big Buddy Lists -- are more valuable). But curiously, it turned out that the agents -- the ''nonloyals'' -- were more effective spreaders of word of mouth than the chain's own fans. Godes and Mayzlin hypothesize that the Rock Bottom's most devoted customers had probably already talked up the restaurant to all the friends and acquaintances that they were likely to tell.

The researchers also looked at the tendency of marketing efforts to focus on ''opinion leaders,'' who often gain that social status by way of expertise. The results here were somewhat mixed, in an interesting way. A loyal opinion leader -- someone who was seen by her social network as an expert on restaurants and who was also a Rock Bottom fan -- was pretty effective; if that restaurant expert was ambivalent about Rock Bottom, she was of little use. In contrast, it didn't really matter if the nonloyal agents knew much about restaurants. What mattered was that they told a lot of people (and presumably that they were enthusiastic). The implication is that it doesn't matter if you know what you're talking about, as long as you are willing to talk a lot.

Godes offered some caveats to that particular conclusion. He pointed out that expertise may be much more important to real-world word of mouth -- the kind that occurs absent an orchestrated effort to create buzz from scratch. He also emphasized that willingness to talk doesn't mean much if you have no one to talk to. Maybe so. But when Dave Balter saw the results, it provided strong evidence for a position he had been coming to for a while: he doesn't quite believe in Magic People anymore. BzzAgent's system does, of course, try to identify who has a large network of friends, who is an expert, who is outspoken, just as Tremor does in its screening. (Actually, several BzzAgents are Tremor members, as well.) ''But we also know that sometimes those people aren't the best at spreading word of mouth,'' Balter said. ''We all get information from people around us who don't fit any type of profile that would make them more intelligent or more focused on products than someone else.'' And the information we share changes, too. ''We might go from influential to noninfluential, from trendsetter to nontrendsetter all year long,'' he suggested, ''because we have continued interactions that change our opinions.''

On some level, then, participating in a voluntary marketing army serves as a kind of consumer-status enabler. You weren't the first on your block with Moon Boots; at least you can be the one to tell your friends about Al Fresco sausage. The more people you can persuade that Al Fresco sausage is good, the better you'll feel about your discovery. BzzAgent, in turn, will help you be a better persuader. Pretty much everyone likes the feeling of having ''the upper hand,'' as Janet Onyenucheya put it. Even in the small orbit of your own social circle, knowing about something first -- telling a friend about a new CD, or discovering a restaurant before anyone else in the office -- is satisfying. Maybe it's altruism, maybe it's a power trip, but influencing other people feels good. As an example of how powerful the desire to have the upper hand can be, consider that some participants in a campaign for a new scent called Ralph Cool simply could not wait for their free sample to arrive and rushed out to buy the $40 product so they could start buzzing. Word-of-mouth marketing leverages not simply the power of the trendsetter but also, as Balter puts it, ''the power of wanting to be a trendsetter.''

BzzAgents are under no obligation to push a product they don't like. In fact, if they think it's awful, they're encouraged to say so. Yet, of all the agents I spoke to, and the hundreds of reports I read, there were hardly any examples of outright dissatisfaction with a product. Most of the agents seemed genuinely excited about most of what they were buzzing.

Part of the reason is that people tend to join campaigns for things that interest them. Perhaps just as important is that the volunteers, hearing that BzzAgent turns down 80 percent of potential clients, seem to believe that the folks at BzzAgent spend their days sorting through the morass of consumer culture, choosing only the best of the best. BzzAgent does want to keep lousy products out of the system, of course, but it also wants to make money. It's a business. And its ability to keep the system relatively free of awful products probably has much less to do with acting as a consumer-culture curator than with the simple fact that there are probably more perfectly good products being sold in America now than at any time in history.

Barry Schwartz, a psychology professor at Swarthmore, is the author of ''The Paradox of Choice,'' a book that addresses the incredible (and at times paralyzing) abundance of options available to the contemporary consumer. In the past, Schwartz notes, the challenge for the consumer was navigating a world of faulty, shoddy or unsafe products. That's not much of an issue anymore. Now, Schwartz told me, Consumer Reports might test 40 stoves, find that 38 of them are pretty good and then resort to sifting among increasingly minor differences to decide which one is the very best value of all, by however narrow a margin. The ''Pretty Good'' Problem complicates our lives as consumers and makes it increasingly difficult for one of those 38 stoves to stand out. But it gives BzzAgent plenty of work.

Still, people's tastes differ, and it seems remarkable that agents are so rarely disappointed. One oddity that Schwartz notes in his book is the ''endowment effect.'' This is one of the many discoveries of the behavioral economists Daniel Kahneman and Amos Tversky. They found that when two items of equal value are handed out randomly to a group of people and those people are given the opportunity to trade, hardly anyone does. It's very unlikely that all the participants were randomly handed the objects they would have preferred had they been asked in advance, so the economists concluded that once something has been given to us, we value it more. In another experiment, conducted in the early 1990's by a psychology professor at the University of Louisville, two groups of subjects were given nine similarly valued objects and asked to rate the desirability of each. The group that was informed in advance it would get to keep one of the items (one of those insulating tubes that keeps canned drinks cold, as it happens) gave that item a more desirable rating than the other objects. The group that didn't get to keep anything rated them all the same. A follow-up experiment found that this ''mere ownership effect'' was essentially instantaneous. Other studies have shown that we like things more simply by virtue of repeated or prolonged exposure to them. (Which could explain why, during the course of that Johnston & Murphy meeting, I gradually went from being indifferent to the shoes to wishing I could get a pair.)

This research on how we value -- or irrationally overvalue -- things that are given to us might help explain why BzzAgents and other word-of-mouth volunteers get excited about whatever they are asked to push. (And if you're curious why, in light of this, you're not crazy about every product you've ever bought, the answer may be adaptation -- our tendency to get used to our possessions and, in effect, fall out of love with them. For the word-of-mouth volunteer this hardly matters, since by the time it happens the campaign is over.) But it doesn't address another mystery: Why would the volunteers work so hard to get other people excited about these products? Another line of research suggests a possible answer. This school of thought would characterize word-of-mouth volunteers as operating not in a traditional money-in-exchange-for-effort ''monetary market,'' but rather in a ''social market.'' A social market is what we engage in when we ask our friends to help us load up the moving van in exchange for pizza. The research suggests that we are likely to get a better effort out of our friends under the social-market scenario than by offering the cash equivalent of the pizza. (A recent article in the journal Psychological Science finds that ''monetizing'' a gift, like the pizza, by announcing how much it is worth, effectively shifts the whole situation from social market to monetary market.) Under some circumstances, we will expend more effort for social rewards than we will for monetary rewards. This suggests that the agents may do more to spread word of mouth precisely because they are not being paid.

Add to all of this the idea that they have been granted status as ''agents'' in an ''elite group'' that most of the world doesn't even know about, and have received a free sample of a brand-new product from a source that they trust, and they are almost certain to expend some kind of effort, unless the product is truly awful.

There is another advantage to the social market. Since the agents are not being paid, and have the option to ignore any Bzz object they don't like, they tend to see themselves as not being involved in marketing at all. Almost all of the BzzAgents I interviewed made this point. ''In marketing, obviously, those people are paid to pump a product, whereas I'm not really getting paid to do this,'' Bollaert, the agent from Brooklyn, explained. ''I don't talk about a product if I don't feel strongly about it. I'll give my honest opinion.''

The notion of the ''honest opinion'' came up again and again in conversations with the agents and with Balter. Seth Godin, the writer and speaker on marketing whose ideas partly inspired BzzAgent, agrees that the agents' honesty is crucial. Paying people to promote products, hiring supermodels to show up in a bar and request a particular vodka, is ''disingenuous, dishonest and almost unethical,'' and it represents a subversion of honest peer-to-peer communication. And honest peer-to-peer communication, he maintains, is the future of marketing.

Godin is not just a BzzAgent fan -- he's also a client. ''Purple Cow'' was marketed through BzzAgent, and Godin quietly plugs the company at the end of the book. He describes BzzAgent as a company at the center of a conversation between its corporate clients and thousands of agents who serve as a kind of guild of consumers. ''I think this is a new kind of media,'' he said. Specifically, this new kind of media is people like Gabriella, or Desjardins, or Bollaert, chatting with friends and strangers.

This argument requires you to accept that a conversation can be honest even if one participant has a hidden agenda. Whether that's possible is something I asked several agents, and Balter himself several times. Of course the agents believe in their own integrity, but that's the easy part. Do we really want a world where every conversation about a product might be secretly tied to a word-of-mouth ''campaign''? Doesn't that kind of undermine, you know, the fabric of social discourse?

''The key is,'' Balter said, ''people already talk about this stuff. They already talk about things they love.'' Manufactured word of mouth is indeed a bad and scary thing, he maintains, but that's not what his company is doing. ''For whatever reason, we have this natural instinct to tell a friend about a product -- and to get them to believe what you believe. We're not trying to change that. All we're trying to do is put some form around it, so it can be measured and understood. That's not changing the social fabric.''

It is certainly easier to defend the voluntary buzz-spreaders as less devious than the paid model pretending to like a product in public -- but the honesty and openness come with an asterisk or two. Those suggestions in the Bzz guides to call bookstores and pretend you don't know the exact title or author you're looking for are pretty hard to define as ''honest.'' Similarly, it's most unlikely that Amazon.com (let alone The Concord Monitor) would consider the reviews of a BzzAgent quite as unbiased and helpful to readers as a review from someone who hadn't consulted talking points compiled with input from the publisher. The whole tone of the Bzz guides -- which read like a cross between a brochure and a training manual -- is a bit difficult to square with the idea of genuineness.

Finally, while BzzAgent tells its volunteers that they are under no obligation to hide their association with the company and its campaigns, the reality is that most of them do hide it most of the time. They don't tell the people they are ''bzzing,'' that they really found out about the sausage, or the perfume, or the shoes, or the book, from some company in Boston that charges six-figure fees to corporations. ''It just seems more natural, when I talk about something, if people don't think I'm trying to push a product,'' Karen Bollaert explained to me. Other agents said the same. Gabriella, for instance, insisted that she really does think Al Fresco makes the best sausage around. Basically, they trust BzzAgent, and they trust themselves, so they don't see a problem.

Nevertheless, Jason Desjardins has told a few people about his efforts for BzzAgent, with mixed results. Some people thought it sounded exciting. Others, however, said they felt ''used.'' One friend he tried to recruit now responds with suspicion when Desjardins talks up something he has done: ''Are you buzzing me?'' the friend will ask. Desjardins shrugs. ''I've been honest about everything.''

One reward Bollaert did collect from BzzAgent was, of all things, the William Gibson novel ''Pattern Recognition'' -- an actual paranoid science-fiction novel about a future in which corporations have become so powerful they can bribe flunkies to infiltrate your life and talk up products. ''It made me think, when somebody says something about a product - I wonder. That gave me a little pause,'' she said. Earlier in our conversation, I touted my iPod. Wouldn't she feel differently about my comments, I asked, if it turned out that I'd gotten it from Apple or a BzzAgent equivalent? ''That's true,'' she said. ''But you know what? If you start questioning everyone's motives, then you'll be in a home with tinfoil on your head.''

The motives of chattering consumers can, of course, be biased in all kinds of ways. If your friend is bragging about his great new cellphone, he may not be a buzz agent, but he may not be the purely rational information source you assume. He says it's the best phone around, and maybe he even believes it -- but the truth may be that he bought it because it looks cool and he read that Jake Gyllenhaal has one just like it. It may be true that we trust our friends more than TV ads, but that doesn't actually mean they've become more reliable.

''I think we all do this naturally anyway,'' Bollaert concluded. ''If you find something you like and somebody asks your advice and you have a product, good or bad, you'll say don't get it or do get it. We're a consumer society.''

Crucial to the BzzAgent system is the small team of young people in Boston who read and answer every single Bzz report. They offer encouragement, tips on how to improve word-of-mouth strategies. Every report is rated and every agent ranked according to a complicated formula, one that is constantly being tweaked, taking into account everything from how often the agent reports to how many people they tend to buzz to the quality of their summaries -- plus intangibles like originality. (This system is part of BzzAgent's defense against people signing up for free stuff and simply making up fake reports about their buzz activities; the home office is trained to spot such things.) Along with the feedback, they almost always throw in a joke or a comment so it's clear that they have actually read the report.

No doubt because of this, many agent reports are full of personality. Some are almost confessional; others are revealing perhaps without intending to be. Casual mentions of boyfriend or girlfriend problems come up, as do complaints about bosses, friends, strangers. One of the most memorable was from a young BzzAgent who reported that a man she met in a bar complimented her on her Ralph Cool perfume, one thing led to another and they spent the night together. The next morning he asked about the perfume again and said he liked it so much he might have to buy some for his wife. (These reports are ultimately handed over to the client -- a trove of anecdotal research from the front lines of consumption.)

Along the way, Agent JonO has become a kind of celebrity, or at least a figure of mystery. There are more calls and e-mail messages and instant messages to ''JonO'' than Jon O'Toole himself can possibly deal with, so lately JonO has become more of a construct than a person. Jason Desjardins sounds honored to have had a chance to meet the real JonO not long ago: O'Toole lined up a dinner in Cambridge with several BzzAgent volunteers, to meet them and hear their thoughts and ideas. Desjardins was so excited about this that at first he overlooked the fact that it was on the same night as his wedding anniversary. (Melissa encouraged him to go anyway. ''But we'll see if JonO is still there for you 10 years from now,'' she said.)

Balter did not count on the agents taking BzzAgent so seriously. He still doesn't seem to know quite what to make of it. He has met only a handful of agents, and while he said he intends to meet more, he sounded almost nervous about it. A number of those he has met have been almost apologetic about not doing more -- about not buzzing enough on this or that campaign. The biggest complaints come from people who say they have not been invited to join enough campaigns. One agent resigned because he said he was unsure whether he could live up to BzzAgent's ethical standards.

This might be the most peculiar thing about BzzAgent: not only are its volunteer agents willing to become shock troops in the marketing revolution, but many of them are flat-out excited about it. At his apartment, Desjardins told me about another book he had read because of BzzAgent. Called ''Join Me,'' it's about a guy who decides he wants to start some sort of voluntary group -- a commune, a cult, whatever you want to call it. He puts an ad in the paper that just says, ''Join me,'' and to his surprise, people are interested. They didn't know what they were joining, or why, but they joined anyway. The guy, whose name is Danny Wallace, decided to turn his followers into a good-deeds army, basically on the ''Pay It Forward'' method. The book is nonfiction.

Why, I asked Desjardins, did people join a group without even knowing what it was? Well, he explained, Wallace's theory was that they just wanted to be part of something. That made sense to me. After all, some people are lucky enough to find meaning and fulfillment through their work, family or spirituality. But many people don't. Many people have boring jobs and indifferent bosses. They feel ignored by politicians. They send e-mail to customer service and no one responds. They get no feedback. It's easy to feel helpless, uncounted, disconnected. Do you think, I asked Desjardins, that there's some element of that going on with BzzAgent?

''I think for some people it probably is,'' he answered. ''For me, it's being part of something big. I think it's such a big thing that's going to shape marketing. To actually be one of the people involved in shaping that is, to me, big.'' That made sense to me too. After all, there is one thing that is even more powerful than the upper hand, more seductive than persuading: believing.

Rob Walker writes the Consumed column for the magazine.