Can One Student Conquer Click Fraud?

9
1454

While bigwigs like Google, Yahoo, and Ask.com join forces with the IAB to form a Vulcan mind-meld and figure the problem of click fraud out, lone entrepreneur/university student Zack Coburn (left) has come up with his own answer. “There are companies right now that are trying to figure out complex algorithms to fix that issue,” he says, “but I think the way is to make an entirely different model for web advertising.”

One way of avoiding click fraud is to sell ads by the number of impressions. “Federated Media is doing that right now, and I think AdBrite has a system for that as well, but even with impressions, you’re opening it up a little bit to fraud,” says Coburn. “It’s not as bad as click fraud, but a publisher could just sit on their page and go to it ten thousand times and make maybe $100 doing that.”

Instead of clicks and impressions, the budding maverick decided it would be better for everyone to start an auction-based ad network that sold advertising space according to time. Coburn’s model in turn resulted in a tiny ad network called Madhens.

Coburn, a member of the class of 2010 at Olin College, wrote up a manifesto pointing out several problems with the online advertising industry. Said statement is accessible from the Madhens site, and addresses problems of click fraud, the lack of control advertisers and publishers have over their advertising, and uneconomical algorithm-based pricing models that benefit the ad network more than the advertiser or the publisher. “As a publisher I had used AdSense and I just saw a number of problems with it,” he wagers, “and those are basically what went into making the manifesto, which turned into Madhens.”

Instead of by clicks or by impressions, advertisers bid on space by the day, by the week, or by the month. It’s something Coburn calls Timeshares, and they’re designed to take away the very possibility of click fraud or impression fraud. “If you click on a publisher’s page, you’ll see how many impressions they’ve served over the past 24 hours, the past week, the past month,” he explains. “You’re allowed to use that as a guideline for whether or not you want to advertise there, but impressions aren’t being used for actually pricing the advertisements. It takes away the incentive for the publishers to click on their own pages.”

Coburn concedes that selling ads by time is not necessarily a novel concept, but he’s quick to point out why Madhens raises the stakes. “There are places like AdBrite that’ll sell you like a seven-day ad for instance, but what makes Madhens unique is that publishers can create a timeshare for any period of time starting now or any time until the future,” he states. Many websites also sell monthly sponsorship opportunities on their own for a flat fee, or for a fee based on the expected number of impressions, but this is the first network this writer’s seen that sells time-based online ads on an auction basis to directly address click fraud.

“It concerns me when there’s no real alternative to click fraud. Google has billions of dollars in revenue from AdSense right now, and what, maybe half of that comes from click fraud?” Coburn wonders. “I think Google will probably come up with some pretty satisfying algorithms in the future, but I’m not sure if the solution is to come up with algorithms, even if you can.”

The auction format optimizes the amount that a publisher can sell ad space for, even if they’re a small publisher with no sales staff. It also lets advertisers know what sites their ads appear on, and it lets them set their price. “Madhens actually lets the advertisers themselves place the bids and see exactly what other people are bidding for the ads,” Coburn further claims. “They can choose where they want to advertise, and then determine how much they’re willing to pay. If you’re an advertiser it’s almost like you’re using eBay. You get to make all the decisions.”

Zack is a programmer by nature. He was only in 7th grade in 2001, when he won the ArsDigita Prize for an interactive website that let computer users shuttle questions and answers back and forth. Coburn’s been active in the online programming community ever since, and in fact, he built Madhens alone. “I thought about naming it Madsense, to let people know it was created out of anger at AdSense for perpetuating click fraud, but that seemed too unprofessional,” he half-jokingly says.

“I’d like it to grow into kind of the alternative for online advertising,” Coburn continues. “Right now it’s serving 10,000 impressions a day, which a small amount. Hopefully it will grow into serving maybe ten times that number.” For now though, Coburn operates Madhens out of his bedroom under the company name “Zeegry.” Zeegry’s site actually lists three employees: Zack, and two others. “Think of the other two as really programs. They’re not real people. Hopefully in college, I’ll find other people who have similar skills to mine.”

Zack Coburn’s approach to advertising certainly is novel. But can something like Madhens, created by a freshman, really solve click fraud, a problem that has vexed companies with billions of dollars in resources? Usually, bigger is better. But sometimes, well, smaller is better too. If the cards fall right, and if Zack’s programming prowess and grandiose thoughts can compensate for a lack of hardened business experience, maybe it can.

9 COMMENTS

  1. I know this is going to sound like super-sour-grapes, but… how does a 19 year old kid who comes up with one semi-original idea about online advertising, get a featured article on ADOTAS?

    We’ve been in business since 2001, we provide eight different mechanisms for publishers and advertisers to collaborate in unique ways, and yet, despite endless announcements and inducements, we’ve gotten VERY LITTLE in the way of coverage like this.

    It’s just so frustrating. My apologies for ranting.

    Jim Raposa
    Founder/President
    Millennium 3 Publishing
    Developer
    SponsorWorks.net

  2. Does anyone really think that impressions fraud won’t be a bigger issue if we move the CPM model? Has anyone even calculated what their CPA would be if they took their current impressions totals from Google and Yahoo and tried to estimate CPM pricing (are you really better off)? As to timesharing, your competitors would find out when your “timeshares” are running and click/cause impressions furiously during those times. I’m not justifying Click Fraud nor saying it isn’t a problem, but none of the solutions I’ve heard in the past few months sound fraud-proof to me.

  3. If you’re going to propose replacing cost-per-click with a new model, try developing something that builds on the value created for advertisers.

    Advertisers have been buying/roadblocking presence since the beginning of time. Putting a new name on it, like “Timesharing” doesn’t change the model. The ability to target by demographics, psychographics and/or intent and to only pay per click is what makes SEM so powerful.

    Of course, this is also what gives fraudsters a platform to work their magic on.

    Want to eradicate click fraud? Launch a network based solely on cost per action/acquisition.

LEAVE A REPLY

Please enter your comment!
Please enter your name here