Barometric is a service that tracks paid media and attributes it to conversions within many different environments. With this knowledge, Barometric has become an industry leader in detecting fraudulent web activity.
With this free tool, publishers can now see the percentage of their website traffic that is considered fraudulent by third party advertising companies. This report will provide peace of mind for your advertisers in knowing that what they pay for is indeed real human traffic.
Traffic Analysis
Last 30 days of fraudulent invalid traffic (IVT) by percentage
0%
Fraud IVT Free
0
Analyzed Pageviews
No Data Available
IVT AnalysisLast 30 days of invalid traffic (IVT) by type
Fraudulent IVT
Benign IVT
No Data Available
Historical Trending
Fraudulent IVT historical and trending analysis
IVT %
Recent Trendline
No Data Available
What is Activity Based Filtration?
Without giving away too much information, the goal of Activity Based Filtration is to find fraudulent pageviews - which in turn trigger fraudulent ad impressions - based on behavior that is non-human.
Good vs Bad bots and spiders...
In most circles, all bot traffic is considered "bad". However, most respectable bots (Google, Bing, Yahoo, etc.) do not artificially inflate pageview counts nor do they request ads. There are a handful of bots that do not care about any of that, and should be considered "bad". Our fraudulent bot metrics include the bad ones, and the good bots are listed under benign IVT.
Benign vs Fraudulent IVT
Fraudulent IVT traffic is typically traffic that is blatantly not traffic that advertisers would not want to serve on. This is traffic that is meant to look legitimate, but is not.
Benign IVT traffic usually comes from valid sources (Facebook, Google, Yahoo, webpage optimizations, etc.) that most ad servers will not display ads on. However, if these benign IVT pageviews are identified correctly by the adserver, then real ad impressions can be generated.
Historical Trending Analysis
Barometric looks at the trailing 12 months of IVT data (if available) to attempt to estimate what the future will hold for the property in question. This analysis involves calculating the mean of the lookback period (the orange line in this graph), as well as machine learning from other properties and pixel traffic throughout the Barometric platform.
Note: Two (2) months worth of data is needed to generate the trend line.