ADOTAS – In June 2013, when my co-founder and I launched MdotLabs to fight the waste to advertisers caused by non-human traffic, most of my contacts in the industry either scoffed at the problem as being something nobody would openly acknowledge or asked jokingly if I thought of myself as some kind of cyber-cop or Dudley Do-Right.
But they were wrong on both fronts; the problem rapidly gained traction as a key issue for the ad ecosystem (thanks in part to lots of press coverage the past few months). And my primary motivation to launch Mdot stemmed from fear and anger.
Why? In mid-2012, as CEO of Broadcast Interactive Media, I had to manage the company through a near-death experience when we were nearly defrauded by unscrupulous publishers and networks.
We’d grown BIM to be one of the top-ranked networks in comScore News and Information rankings and, as we sought to grow beyond local news, on-boarded several publishers we thought were high quality. Some were even funded by top-name Silicon Valley VCs. BIM had checks in place to ensure quality (manually checking privacy policies and sites, viewability tools, brand-safety and content scoring, etc.). But none of these protected us from what happened next.
We were running most of our publisher inventory (hundreds of sites) through Google’s DoubleClick AdExchange. Suddenly, Google hit the BIM network with a large clawback (as they did to so many in 2012) and of course gave us no data on what they’d detected or on what specific sites. A huge revenue deduction was magnified as we not only wouldn’t get paid by Google, but we also we had no data on which of the hundreds of sites we worked with were responsible for the issue. So we didn’t know who to withhold OUR payment from and which publishers were good and should still get paid.
Fortunately, we did have a small team of data scientists that went into overdrive to review our own ad serving and hosting records, and my co-founder Paul Barford was able to apply anomaly detection techniques he’d perfected in the cyber-security domain to determine the likely offenders. It became clear we’d been hit by sophisticated traffic generation networks using botnets to mimic human activity. But we couldn’t have seen it coming given then-current detection tools in the market, as the bots rendered the ads “viewable” with browsers and ran the ads on “safe” content.
After many rounds of legal letters and threats of litigation (and many sleepless nights on my part), BIM was able to prove which sites committed the fraud and withhold payment from them. What we launched while in crisis mode to protect ourselves and honed for several months became the platform we’d spin off into its own company.
Stories like mine aren’t unique. Many of us have suffered sleepless nights at least because of ad fraud, others have lost much more, and it’s important to remember that as each of us in the industry weigh the consequences of turning a blind eye.
MdotLabs is a data security platform that prevents the impact of fraudulent ad traffic, and growing a successful business is great. But my personal motivation for starting this company is that no entrepreneur should have to go through the painful experience I did when we were attacked at BIM.