08 Aug Online Political Advertising: The Prehistoric Era Continues
Several types of online communication have enjoyed break-out moments in U.S. political campaigning and public affairs, including e-mail (Jesse Ventura 1998), fundraising (McCain 2000), blogging (the Trent Lott resignation 2001), the organization of in-person meetings (MeetUp/Howard Dean 2003), and web videos (the “Macaca” incident, 2006). After these successes received notice, adoption of the practices behind them spread among campaigners and activists. They became standard equipment in the online politics toolkit.
Online advertising would seem a likely candidate for this social treatment. As Henry Copeland and Megan Mitzel of Blogads.com point out, the interactivity, accountability, iterability, and targeting capacities of online ads -not to mention their relative low cost-make them an attractive complement to campaign advertising in print, broadcast, and cable media. Yet despite steady growth in overall online advertising, at a rate of roughly 1% additional share of total advertising spending per year since the dawn of the millennium, and steady if not comparable growth in online political advertising expenditures at the presidential level, there has been no break-out moment of social discovery and adoption.
In this essay we examine the state of the craft of online political advertising in the 2004 and (2007 phase of the) 2008 presidential campaigns. We contend that online political advertising remains in a prehistoric era. It exists. It is maturing in sophistication of strategy and message. But it lacks a killer application and good public metrics. To borrow an image and sound from one of the most famous scenes in film history, the ape has not tossed the bone into the air to the fanfare from “Also Sprach Zarathustra. ”
2004: Banner Ads and Ambient Persuasion
Our 2004 analysis begins by drawing from a report written for the Pew Internet & American Life Project; it relies on data from Evaliant Media Resources, an affiliate company of TNSMI/Campaign Media Analysis Group. Evaliant used “spidering” technology to search thousands of web sites for brand-related banner advertising. Banner ads fall into the category of online advertising known as display. Along with such other display templates as skyscrapers and rectangles, the banner ad is a purchased space on a web site intended to be viewed by Internet users who have come to the web site to see other content and do other things. Display advertising online, like street billboards, yard signs, broadcast commercials, and print ads, interrupts the gaze of people to make an impression on them. Evaliant collected banner ads, coded them according to their site locations (per page) and daily frequency, and then estimated the price paid for each perceived exposure based on available rate information.
In the first eight months of 2004, the time period for which we have data, the presidential candidate campaigns, national parties, and major 527 advocacy organizations spent an estimated $2.66 million on banner ads. This amounted to less than 1% of the buy for television ads in the top 100 markets during the same time period. The Kerry campaign alone raised twice as much money as this total online in one day: $5.7 million on July 29, when he gave his speech accepting the Democratic nomination for president.
In placing their banner ads, the two presidential campaigns appeared to prefer local to national and global news outlets, and web sites of traditional media properties to those of online companies. It’s unknown whether the buys at national and global outlets were targeted to particular segments. The Bush campaign aimed its online advertising at middle-class women, Hispanics, and voters in battleground states in one big blast in May, which cost approximately $400,000. Its top five ad buys were at KPTV Oregons12.tv.com (the Fox network affiliate in Portland OR), Parents.com (Parents magazine), KNVA-TV.com (WB network in Austin TX), ElNuevoHerald.com (Miami FL), and KPHO CBS 5 News.com (Phoenix AZ). The Kerry campaign concentrated on raising money from progressive outlets in metropolitan areas. Its top five ad buys were at SFGate.com (Chronicle newspaper, San Francisco CA), Newsweek.com, VillageVoice.com, Reuters.com, and L.A. WeeklyMedia.com.
An examination of 137 display ads in the archive attributable to the Bush and Kerry campaigns and the Democratic and Republican National Committees (DNC and RNC) between January and July 2004 reveals mostly slogans and graphics of the kind found on bumper-stickers and billboards, although the text were longer. Forays into flash animation were rare. No ad announced a political endorsement, issued an invitation to rallies and meetings, or referenced upcoming events such as primaries, television appearances, and financial disclosure deadlines. While the display ads were targeted, the contents were, on the whole, generic. Only one ad asked viewers to click through to a specific message: a 2 ½ minute video featuring Laura Bush talking about education. There was no discernable pattern regarding the use of negative, contrast, and positive ads.
We know something about the strategies and results behind some of these ad buys thanks to an article by Michael Bassik. The DNC, the Kerry campaign, and the consulting firm Malchow Schlackman Hoppey & Cooper (where Bassik works) collaborated on several banner ad purchases. More interestingly, they ventured into an exercise in what might be termed ambient persuasion.
The goal was to shape public opinion about the outcome of the first presidential debate between Bush and Kerry. The method: in the hours just after the debate concluded, strew display ads on more than fifty Web sites where Internet users were likely to soak up news accounts of the debate, including the home pages of the sites operated by Reuters, The New York Times, the Washington Post, USA Today, MSNBC, and even The Weather Channel. The banners read like headlines: “Debate Shows Kerry’s Strength -Bush Fails To deliver Plan for Iraq;” “Kerry Presents Strong Plan for America -Bush Won’t Admit Failures.” The mind’s eye would presumably gloss over the distinction between ad and news content, and thereby be inclined to take the former with the credibility of the latter. A study commissioned by the collaborative found that 55 % of those who saw the ads thought Kerry won, compared with 49% of a similar survey group who did not see them. As a bonus, more than $1 million in contributions flowed into the DNC via the “Debate Center” landing page reached by viewers who clicked on the banners. The DNC sustained the online advertising campaign until the election.
2008: Early Patterns and Prospects
The online world had changed in several respects by the time the next presidential campaigns began to advertise. Video portals and social networking sites were providing new forms and forums for content. There were more companies offering data on advertising results, including ads on search engines. As in 2004, a majority of web users (52%) told interviewers that they encountered campaign news and information while online doing something else. But the percentage of Americans saying they learned something about the campaigns through the Internet almost doubled between 2004 and 2008 (13 to 24%). More voters had broadband connections, and they had more experience as consumers and producers of online political information. Clearly, the opportunities for online political advertising had improved.
Among the large number of entrants in the presidential nomination races, three stood out in 2007 for spending on Internet display advertising: Republicans Mitt Romney and John McCain, and Democrat Barack Obama. Combined, the campaigns for Romney (37%) and McCain (34%) ran over 70 percent of the online display ads purchased by the candidates between January 1 and December 16, 2007, according to data from Nielsen Online AdRelevance. Obama’s campaign ran more than a quarter (27%) of all presidential campaign ads in that time, beginning in the summer. Display ads for Hillary Clinton, Mike Huckabee, and the others accounted for just 1 percent. All told, the candidates ran more than 277 million display ad impressions. (An ad impression marks one view by a web user.)
While we have a relatively solid estimate of which political campaigns, particularly presidential candidate campaigns, ran online display advertising in 2007, some ads running on small sites such as blogs and ads that were not correctly categorized may have fallen through the cracks. Also, when it comes to tracking other types of online advertising formats, there is no public data tracking ads in any systematic manner. So we only have anecdotal information and interview material to ground our analysis of non-display political advertising formats, including in-stream video advertising (whereby spots play before a requested clip on a local news site, YouTube, or elsewhere), and text ads placed on the return pages supplied by Google, Yahoo, MSN and other search engines according to user requests.
Display Ad Content Broadens
In contrast to 2004, display ads in 2007 featured a wider array of messages and purposes. Fundraising in small donations remained a popular approach, as it was in 2004; Mitt Romney asked Web viewers to “Donate $44 for the future 44th President.” But other ads run early in the year by the Romney for President campaign pushed local campaign events, urging people to “Join the Rally Today.” Many of Barack Obama’s display ads served later in the year were also intended to get people to attend campaign events, perhaps banking on curiosity about the candidate’s famed oratorical skills and glamorous presence: “Get to Know Barack Obama. Attend Invitations to Campaign Events” and “Show your Support. Attend Local Campaign Events.” Most of the Illinois Senator’s ads culminated with a plea to “Join Us.”
Before his candidacy became official, the John McCain 2008 Exploratory Committee ran video ads on AOL in conjunction with the interactive technology firm PointRoll. “Be There From the Start,” read a small display ad running on the conservative community, news, and opinion site Townhall.com, as well as on a collection of sites selected via Google’s AdSense service. The McCain ads featured invitations to view three thirty-second videos presenting the potential candidate. Users submitted contact information as they clicked through the ad to the videos, including physical and e-mail addresses. After the viewings, they could link to a donation page on the pre-campaign site. The advertisements thus simultaneously tested messages, acquired data on prospective supporters, raised funds, and promoted the idea of the candidacy.
McCain also gathered grassroots information by promoting online petitions and surveys. “Surrender is not an option,” declared one ad, asking people to “Sign the Petition Today” in a show of support for the U.S. military mission in Iraq. McCain was not alone with this technique at this time over this issue. A display ad run by the Hillary Clinton campaign called upon its viewers to “Sign the Petition” expressing opposition to President Bush’s veto threat against Iraq War-related legislation. “Sign the petition to stop the Bush veto of the will of the people. Start bringing our troops home,” read another such ad, which was placed on CNN’s News and Politics sections. A third version told Web users that “Your signature can be as powerful as President Bush’s. Tell Bush: listen to the people on Iraq. Sign the Petition. End the Veto threat.”
When it came to the issue of Congressional (over-)spending, McCain’s online ads adopted a lighter tone. One that ran for several months showed googly-eyed cartoon nuts and citrus fruit leaving the Capitol. “$74 million tax dollars for peanut storage costs? That’s Nuts! $100 million tax dollars for citrus assistance? Orange You Outraged?” On it went, flipping to salmon and, of course, pork.
In June of 2007, Republican presidential hopeful Congressman Tom Tancredo briefly ran display ads on The Drudge Report news portal focusing on his signature issue, immigration. The ads were timed to coincide with a Senate vote on an immigration reform bill that, in his view, offered amnesty to illegal immigrants. A petition was attached so that electronic signatories would declare “Here is my message for any politician who supports an amnesty bill: I will commit myself to working for your defeat!” The year before, grassroots activists employed the Internet to stop a similar bill in the face of bipartisan Congressional and White House backing.
Branding, the marketing term that represents the experiences associated with a good or service, was an evident purpose behind the display ads of lesser-known candidates. In political terms, the purpose of the ads might be best categorized as name recognition or persuasion. In one ad placed days before the Iowa caucuses, Democratic Governor Bill Richardson directed viewers of the web site of the Ackley World Journal to “Read Bill Richardson’s plan for behavioral health care” on his web site. The ad summarized the plan so as to depict Richardson as a details-oriented pro-government services candidate: “A Heroes Health Card for veterans, ensuring them the quality care they deserve. Quality health care for PTSD [Post-Traumatic Stress Disorder] and other mental trauma. Access to affordable behavioral health care for all Americans.” To a degree, this was more a re-branding than an introductory stamp, given Richardson’s series of mock job interview television and web spots in which he humorously advanced an image as the candidate with the best qualifications for the presidency.
Attack ads seemed sharper than in the previous campaign cycle. In ads served in October 2007 on FoxNews.com and such conservative blogs as Power Line, Biblical Womanhood and Hugh Hewitt’s blog on Townhall.com, Republican Fred Thompson’s campaign urged readers to “Support the Real Conservative.” The ads promoted the brand by taking aim at both Romney, questioning the his fidelity to a pro-life stance, and Rudy Giuliani’s pro-choice position. The ads suggested that these Thompson rivals were feigning conservative values to garner support from important far-right Republican primary voters. “This is not a time for philosophical flexibility, it is a time to stand up for what we believe in,” noted the animated display ads, which linked to the Fred08.com campaign contribution page. The effort came just in time for a FOX News debate among GOP presidential hopefuls and a Values Voter Summit hosted in Washington, D.C. by pro-marriage and family group Family Research Council.
In what appeared to be the first use of video overlay advertising by a presidential campaign, Romney ads reprised –or, in online lingo, “repurposed”– television spots running concurrently on the air in Iowa. Web users watching videos related to family life would suddenly see an image overlaid on top of the video inviting them to watch a Romney ad about the same subject. (The “Our Home” ad combined reality-TV style footage of the former Massachusetts Governor bouncing grandkids on his knee with family film archives of his wife, Ann, with their kids.) Clicking on the overlaid image would start the ad video, pausing the originally selected video clip. The Romney campaign bought time for these overlay ads through the technology firm ScanScout, targeting the ads based on socially conservative and family-related keywords identified in the audio tracks of the Web videos. This type of ad placement is known as contextual targeting, because it keys off the subject matter of editorial content.
The Romney campaign and others also ran display ads on blogs. Blog ads can serve the purposes already discussed in this section while advancing two additional strategic goals: the goodwill of influential bloggers (some of whom depend heavily on ads to sustain their chosen vocation of contemporary pamphleteering), and rapid responses to charges that bloggers discuss in real time. Blog ads are inexpensive even by Internet standards: a top slot on DailyKos, attracting more than six million impressions a week, costs $15,000.
Ad Networks: Smart Placements…and Placements That Smarted
Ad networks offer placements in a multitude of web sites according to targeting criteria. They are a rational development, a huge time-saver given the gigantic number of possible locations to reach web users. These ad networks offer choices not just of specific sites, but of web pages, times of the day and week, and designated groupings of users by demographic, geographic, and behavioral categories. (Behavioral targeting entails delivery of ads or other web content to computers whose users have previously viewed certain editorial content, clicked on an ad, or taken another such detectable action.) Hundreds of online ad networks exist, from Google’s fortune-making AdSense to ValueClick and Advertising.com.
In 2007, the presidential campaigns ran the bulk of their ads on the big-name portals and news media sites including FoxNews.com, The New York Times, MSNBC, Newsmax and HuffingtonPost.com. According to AdRelevance, the top site for presidential campaign ads was Yahoo!, which ran 32% of the ads the monitoring operation identified. The ads ran on news and politics pages, but also in Yahoo!’s movies, sports, and e-mail sections. MSN served about 11%, Excite 6%, and AOL 4%.
Some ads may have appeared as well on the “long tail” of the distribution curve, where traffic can number in the hundreds or even dozens of impressions. Though we lack the data to substantiate that phenomenon, AdRelevance data showed some online ad network buys helped push political ad dollars out to niche content sites. The three top display ad spenders, Romney, McCain, and Obama, had ads show up in unlikely Web nooks and crannies. McCain ads appeared on RealityTVWorld.com, Southern Living Online, and HowStuffWorks, while ads for Obama ran on GoComics, Hoover’s Online, and CNET TV.com. Romney for President ads were also seen on the gaming community site Allakhazam’s Magical Realm.
Since the ad network placement process is abstracted through the selection of key words or audience demographic categories, and often relies on automation, ads can show up alongside inappropriate content. Buying through an ad network embarrassed conservative Mitt Romney when his campaign ads surfaced on Gay.com, Advocate.com, and PlanetOut.
Social Networking Sites: Hot Spots of the 2008 Cycle
Candidates have flocked to establish presences on social networking sites, especially MySpace and Facebook, which have experienced astronomical growth in participation during the past few years. Most presidential campaign web home pages sport iconic links to social networking sites, a sharp break from the tradition of not putting up any exit gates to tempt visitors. Some candidates and their family members (especially those in the 18-30 age group) have put up profile pages, videos, and blog entries on social networking sites, and their campaigns have worked with members to promote the activity which gives this type of site its name. But with some exceptions, most campaigns did not buy ads on these sites. Democrat John Edwards did purchase approximately 170,000 impressions on MySpace in October 2007, according to AdRelevance. It was a simple, single-image ad with the caption “Join the campaign to change America. Join Senator John Edwards at http://www.JohnEdwards.com.” McCain ads also showed up on MySpace during the primary season.
How to explain this combination of reluctance and enthusiasm? The fact is commercial and political advertisers alike are concerned about having their ads appear alongside offensive user-generated content. We know that, in its deal with social network site publisher Community Connect, Obama’s campaign sought to minimize the risks of ads served by ad networks, by obtaining the assurance that only house banners promoting Community Connect itself would be posted on Obama profile pages. The goal was to prevent inappropriate ads from showing up on those profile pages. For instance, other profiles on AsianAve.com might show text ads served by Google promoting “Asian Girl Photos.”
The publisher also agreed to run homepage display ads to drive traffic to those pages on its family of networking sites, which span several identity cultures. (In addition to AsianAve.com, Community Connect owns BlackPlanet.com, the gay community network site Glee.com, Latino-oriented MiGente.com, and the Christian networking site FaithBase.com.) Community Connect also sent out e-mail alerts to site members notifying them about the official Obama presence. Yet as for a full-fledged paid ad campaign on the network’s sites, that was not part of the deal; a Community Connect representative expressed the hope that advertising would come in time. The Clinton and Edwards campaigns subsequently contacted the publisher in efforts to obtain similar deals.
Meanwhile, as one might expect, the Obama campaign tailored the content added to each site. Obama’s Glee.com profile included a post about National Coming Out Day, while his MiGente page touted the endorsement of his energy plan by former U.S. Secretary of Energy Federico Pena. The candidate’s profiles on AsianAve.com, MiGente and BlackPlanet.com spotlighted education and family issues, while statements on Glee.com highlighted environmental issues. FaithBase.com and AsianAve.com noted Obama’s thoughts on faith and politics. The candidate’s main “personal message” on all the sites was the same, noting, “I was fortunate to be able to grow up seeing America from varied viewpoints.”
Search Engine Ads
Sponsored links or search ads are purchased directly from the companies that design and operate the engines powering this near-universal online activity. The ads appear according to how relevant they are to a user’s search request, or how much the advertiser is willing to pay for a particular slot, or a negotiated combination of the two criteria. They are usually text-based, although Google’s AdSense and a few others offer programs whereby image-based ads can be targeted according to keywords users search for. Most search advertising is sold on a cost-per-click or cost-per-action basis, meaning that the advertiser pays only to the extent a viewer clicks on the ad or performs a “conversion” activity such as signing up for a newsletter.
Click-throughs and conversions have become widely accepted metrics of success in the online world. But there are no measurement services that comprehensively disclose who is paying how much for what kind of search ads with whom, on what sites, with what results. A report by the search engine marketing firm iCrossing, published in July 2007, stated that presidential campaign involvement in this major activity remained sparse. To make its assessments, iCrossing analyzed 126 election-related issue keywords along with candidate site URLs. The company also factored in rank and coverage data from search marketing research firm AdGooroo and Google AdWords data on keyword costs. In addition to estimating search ad spending and visibility, it commissioned Opinion Research Corporation to survey more than 1,000 Web users concerning their search habits when it comes to election-related information.
In May 2007, McCain’s presidential campaign spent less than half than the Edwards campaign did on issue-based search ads in May, but McCain got far more bang for his buck. McCain came out ahead in visibility, appearing prominently in searches for “stem cell research,” “pro-life,” “campaign finance,” “electoral reform,” “ethics reform,” “government accountability,” “government reform,” “lobbyist,” “special interests,” “tort reform,” “DNC,” and “RNC.” Edwards’s ads were highly visible only in results for searches on “Iraq” and “war in Iraq.” Romney also expended funds on search ads: 5% of the detected and calculated total, according to the report, compared with Edwards’s 64% share and McCain’s 29%. Ads for the former Massachusetts governor surfaced in search results for “ethics,” “family values,” “war in Iraq,” and “social conservative.” Obama’s campaign accounted for 4% of spending, with sponsored links appearing in searches for “Iraq” and “war in Iraq.” Giuliani showed up in searches on “flat tax;” and Paul for “war in Iraq” queries.
The report’s survey found that 89% of voters using search engines to track down election information have conducted searches on a relevant issue. Most searches followed party lines. Obama was the most-searched candidate of all, prompting searches by more than 50% of all people using search engines for election information, and 60 percent of the Democrats in that group. Forty percent of all election searchers sought information on Clinton, and about 57% of Democrats in that segment did. Thirty-seven percent searched for Giuliani, while 51% of Republicans in that segment searched for the Republican candidate. About 23 percent of all searched for McCain, and 28% of Republicans did.
Independents searched for lesser-known and potential candidates more than the self-identified partisans. Republican Newt Gingrich was sought by 18% of Independents, and just 10% of Republicans. Huckabee (a second-tier candidate at the time of the survey) drew searches from less than 5% of Republicans compared to 12% of Independents. Kucinich was searched on by just 5% of Democrats, but drew searches by 14% of Independents.
The online consulting firm Connell Donatelli, which placed the search ads for McCain, claimed the campaign has reaped $4 in fundraising cash for every $1 spent on search advertising. As of July 2007, search had accounted for around 40 percent of the campaign’s online donations. The campaign also ran display and video ads through Google’s AdSense network, which places keyword-targeted text and display ads on non-Google web sites.
Social adoption of a new technology or technique often soars upon demonstration of its superiority, repeatability, and verifiability. The new mousetrap, in other words, must not only be seen as better, it must be better when prospective adopters try it themselves. When these and related criteria are met, a community embraces a killer application: a utility so compelling that in order to take advantage of it people will purchase new equipment and incorporate it into their lives. (When it sits on the shelf, it doesn’t count as adopted.) As we have seen, the internet provides a host of advertising formats, methods, and measurements. What might be a winning combination capable of providing online political advertising with its break-out moment?
Finding the right combination will be tougher than it looks. While measurements abound in the world of the Internet (which is, after all, a network of computers), statistics alone are insufficient persuaders of value. Head to the online politics news site TechPresident.com, and you will see a “Charts” bar. Each tab on the bar leads to interactive graphs with brightly colored data points detailing the number of the presidential candidates’ MySpace Friends, YouTube Views, Facebook Supporters, Technorati Tracks, Eventful Demands, Hitwise Traffic, and Meetup Members. The bar reinforces our argument at the elemental level: there is no tab with statistics about ads. It also helps us make a subtler point: These statistics are fun to ponder, but hard to use, because a formula to derive value from the data has not been devised, demonstrated, and accepted. They are not yet success metrics.
To professional politicians the worth of online metrics, as with poll data, depends on their reliability as indicia of expected dollars and votes, the supreme measures of value to campaigners. To political scientists and other students of democratic politics, the most valuable correlations are with voter knowledge and participation. So, for example, most professional pollsters regard the “is this country going in the right direction or headed down the wrong track” question as valuable because they have a good idea, if not a solid calculating formula, of how well responses to that question correlate with the electoral prospects of incumbents and challengers.
The most obvious candidate for a valuable metric of measuring the state of online political advertising is spending. When spending rises considerably, that may be a sign that more professionals appreciate what online ads can do for their campaigns. Reports released in late 2007 and January 2008 indicate the lack of a clear gauge of online political ad spending. Late in 2007, the market research firm PQ Media forecast a 150% increase in spending by political campaigners on Internet advertising, marketing and promotional efforts. However, the estimated $73 million expected to go towards the web in 2008 represents a 1.6 percent sliver of the projected campaign media spending pie. And, according to PQ, the majority of dollars spent by political advertisers on the web will go towards email marketing efforts, a primary method for online fundraising pitches, rather than for display, search or video advertising. Other reports released in January 2008, sometimes including public policy advertising in addition to candidate and advocacy group campaigns, suggest online ad spending by political advertisers will total anywhere from $20 million to $110 million in 2008, a huge discrepancy caused by varying methodologies.
Yet given the relative efficiency of internet communication compared with other media (including transportation costs for face-to-face meetings), raw spending is not the best indicator of value. A better approach to measurement is already employed by commercial advertisers, who consider a variety of metrics to determine ROI, or return on investment. In fact, all advertisers can measure ROI for online ads much more directly than they can for broadcast and print ads, and even more than for the impact of such online assets as having a MySpace page.
ROI is a familiar concept among commercial advertisers, but foreign in the political sphere, where returns connote election returns, which come once or twice every two years, at most. And, as we have shown in our review of current practices, ROI on online ad spending encompasses multiple types of returns: informational signups, name recognition, opinion change, rally attendance, and, of course, donations and donors. How should these returns be combined into a valuable formula? And who can be trusted with the often proprietary data to plug in the numbers?
Commercial ad veterans like Eric Frenchman of McCain’s consulting firm Connell Donatelli already rely on ROI-based metrics. For example, in the case of search ads, John McCain 2008 compared the amount of fundraising dollars collected after someone clicked on a search ad with the amount spent on the ad to determine ROI. Mitt Romney’s online campaign staff measured ad success by the number of “Team Mitt” volunteer signups gathered as a result of ad click-throughs, in addition to contributions collected after supporters clicked on the ads, according to Director of eStrategy Mindy Finn. Another way the campaign measured online advertising was by devising a formula based on the number of potential voters in a particular region or state reached per dollar spent on web ads.
We hope that effective and standardized metrics for measuring the impact of online political advertising are not long off in coming. Because online ads can readily elicit responses from viewers, they have the potential to merge the professional values of campaign money and votes (and the secondary value of journalistic attention) with the idealistic values of voter knowledge and participation. The killer application, in other words, could be a rise in grassroots activity generated per dollar invested. Like any political practice and business enterprise, online political advertising will have its greedy and venal side. But the civic benefits are not just theoretical wishes; they are already evident in the scattershot and sporadic results we have seen.
Someone should pick up that bone and figure out what to do with it.
Michael Cornfield is VP for Research and Media Strategy at 720 Strategies, a Washington D.C. based public persuasion firm. Kate Kaye is a reporter and editor for ClickZ News, a daily trade publication covering the interactive advertising industry.