How to Bust Online Marketers Fraud

How to Bust Online Marketers Fraud

by Asaf Greiner, Nov 27, 2017
 

Asaf Greiner, Protected Media Founder and CEO warns us against fraudulent ads. Despite all the regulations, standards and consortium’s, ad fraud is still siphoning of huge portions of advertisers’ budgets. Today amateurs, with no programming experience can purchase or rent bots, and install them all over the web. Fraudsters can even cover their tracks by manipulating ad measurements to keep under the radar

Beating online ad fraud, like any other process for solving a complex problem, is an activity that involves constant analysis and learning from previous experiences. Having good clean traffic is a result of good habits. It’s an iterative process where based on results, the plan, parameters, and traffic samples are regularly adjusted. This article recommends a disciplined approach with a goal, plan of action, specific data parameters, and corrective measures to bust ad fraud. Here are some tips for you to protect from fraudulent. 

  • Create program goals: It’s important to state up front the goal of the quality traffic program.  Are you concerned that your ads will appear on sites with inappropriate content such as porn, tobacco, firearms, extremist views, etc.? Are you working with a new exchange or ad network and you are concerned about the quality of their traffic? Are you attempting to meet quality standards set by a specific regulatory body or media agency? It’s also important to decide what type of corrective action you are prepared to take if an alarm goes off when fraud is detected. Will you ask for a refund or just eliminate the traffic source all together? If you are working with a larger exchange, will you cancel out the bad sources while keeping the good ones? This decision will impact the type of reporting you need.
  • Select relevant data: You need to design reports that include actionable data that will enable you to follow up on your plan. Typical parameters that set off an alarm can be unrealistic click-through rates, bounce rates and page duration times, and if the IP address of a clicker is a server or data center or appears on a blacklist. If the plan is to ask for a refund for bad traffic, you may need to measure performance per publisher, site, page, placement, geo, time of day etc. The more precise the measurements, the higher the likelihood that you can pinpoint the problematic source and receive a refund, or cancel out the exact source of bad traffic.
  • Build a sample: Verifying all the traffic isn’t always economically feasible. Sampling micro-sets of ad traffic can reveal the trickiest of scams. Analyzing a random subset of all data will uncover key clues in the larger data set, keeping the overall cost of inspection more affordable. So, for example, with a 1% margin of error and a 99% confidence level, one could analyze how many fake coins are in a shipment of 1,000,000 gold coins and do not have the proper weight. To do this, you only need to randomly sample 16,641 coins, reducing inspection costs significantly. By cutting down on the number of inspections you could allocate more resources to each inspection, increasing the likelihood that fraud will be detected.
  • Take corrective action: When you find fraudulent traffic, you can suspend or disable invalid accounts, and withhold payments. You can receive money back or have your account credited not only for the month where you found the invalid activity, but for previous months as well. Having the fake clicks and views well documented is important for the negotiation process.
  • Adjust media buying behavior: Your media buying plan should be modified based on lessons learned about fake traffic. It’s possible based on the data to build a profile of suspicious traffic sources that are likely to include fraud and then take them off your list. It’s all about predicting and then avoiding fraud.

to the full article