Mobile ad network teams up with academics to safeguard ad quality

BuzzCity announces joint research initiative to improve advertiser results
  BuzzCity announced today that it is intensifying steps to safeguard ad quality across mobile advertising networks, by undertaking a study with Computer Science researchers from the National University of Singapore (NUS).
In addition certain countries are experiencing triple-digit growth including Kenya, Pakistan and Argentina. BuzzCity has seen tremendous growth across its mobile advertising network with a 95% increase against Q1 of 2011. In this first quarter, more than 45 billion ad banners were delivered across the entire network of more than 10,000 publisher sites, reaching an average of over 300 million unique users per month.
CPC (cost per click) is the dominant model driving advertising activity on mobile ad networks and search. The way it works is whenever a site visitor clicks on an ad, a small amount of money moves from an advertiser account to the site owner’s account. Advertisers seek out networks that offer them the best return on their investment. The industry speaks of finding ‘traffic that converts’ – clicks that result in the advertisers’ goal of a sale, a signup or a download.
Problems can arise when fraudulent site owners generate clicks on ads that appear on their own site to boost their earnings. They might simply click on the ads whenever they get a chance, or they could try to alter the codes in an ad tag, or they could even use ‘bots’ -- software designed to click on banners. Such ‘click fraud’ seriously diminishes the efficiency of a campaign, and ad networks invest significant time and energy to stamp out the practice. It can also damage user experience – with promotional messages becoming unreliable and misleading.
‘As mobile traffic and internet advertising grows, click fraud has the potential to diminish the trust of advertisers who otherwise should be getting better returns from their growing mobile budgets,’ said BuzzCity CEO, Dr KF Lai. ‘BuzzCity already has a system in place to detect and discard fraudulent clicks before they reach advertisers’ accounts, but we want to ensure that it doesn’t become an issue moving forward.’
BuzzCity’s sophisticated algorithms look for specific patterns in the data generated. This includes IP addresses, user sessions, carrier information, and browser detection, all of which provide clues that help BuzzCity’s systems to monitor potential fraud.
Their collaboration with a research team at the National University of Singapore will take this further, and follows the recent appointment of Dr Lai as an Adjunct Associate Professor with the Department of Computer Science at NUS. Researchers in Department of Computer Science will use their expertise in informatics to analyse the volume of clickstream data that will be provided by BuzzCity.
‘While it may never be possible to reveal the intent behind a click, research like this – whether it is a competitor who has a vested interest in draining its rival’s ad spend, or a publisher who wants to earn more money without investing in the audience – will help create mathematical models to better identify quality traffic for advertisers, and improve earnings for publishers,’ said Provost Chair Professor and Head of Computer Science Professor Wong Lim Soon of the National University of Singapore’s School of Computing.