The war between fraudsters and anti-fraud solutions shows no signs of slowing. But that doesn't mean we should give up. Here we will explore some features Adserver.Online has to offer to help you fight against traffic fraud.
- We do NOT guarantee we will filter out fraud/bot traffic completely
- We do NOT guarantee we will NOT filter out real human visitors
Filters configuration is in the Settings/Anti-fraud menu and offers the following options.
Bad User Agents
This filter allows blocking requests from non-standard browsers or utilities like Wget or Curl. Unsophisticated fraudsters can use those.
Usually, traffic generation software or scripts are hosted on different data centers (i.e., Amazon). This filter allows filtering out requests from data-centres' IP ranges.
Fraudsters may use proxy servers or VPN services to fake their geo-locations, including Tor. This filter allows detecting requests from IPs that belong to proxies and VPN providers.
Please note this option won't filter out peer-to-peer VPN services like Hola. They don't use any actual servers. Therefore, they cannot be blocked. Hola users who installed their browser's extension become servers. And Hola uses these users to proxy requests from other users.
Putting ad code into small or invisible IFrames is a favorite fraudster strategy. This filter tries to detect code running in iFrames and prevents stats counting.
Please note you should NOT use this filter if you are going to re-publish our ad code on other ad networks. 3rd-party ad networks would put our code into IFrames, so our code would not work correctly.
3rd-party Traffic Validation Providers
Besides build-in filters, we support the following 3rd-party integrations:
In the following article, you can learn how to configure traffic validation providers.
Be aware that 3rd-party filters only apply to clicks (banners and similar format, VAST) and to redirection-based formats (direct links, popups). Hence, they don't apply to impressions, views, video events, or conversion tracking.
Note 3rd-party filters do not replace built-in filters. Instead, they work in parallel with them.
Filtered Traffic Counting
Filtered-out traffic will still add up to the total number of billable ad requests, as our servers will have to handle these requests in the usual way.
By default, anti-fraud filters don't affect the ad serving itself – everything will work as usual, but the stats will not reflect detected fraud/bot traffic.
You may tick the "Forced blocking" checkbox if you don't want to show ads to fraudsters. In this case, 0.5 billable ad requests will be counted.
In the stats report, you may enable the "TrafQ, %" column that stands for Traffic Quality – a higher percentage means higher traffic quality. Note this calculation is based on built-in filters only.
Fraudulent Traffic Ignoring
It's not a secret that publishers may provide low-quality traffic or (try to) cheat to increase their revenue. Unfortunately, it's become a part of our life.
To mitigate the effect of fraudulent activities and low-quality traffic, we have implemented a special feature that allows account owners to keep their revenue under better control by "ignoring" a set percentage of the publisher's traffic. In order to do that, you should specify the default percentage of clicks and impressions to be ignored in the corresponding fields in the Settings / Anti-fraud menu.
Our ad server processes advertisers' and publishers' stats separately; this feature only affects publishers' stats. Their stats (in owner's and their own panels) and revenue will be reduced according to these settings. At the same time, advertisers' stats remain the same.
For example, if you set the zone's fraudulent impressions to 10%, then ~10% of all impressions of that particular ad zone won't be added to the publisher's stats.
These settings do not affect advertisers' stats in any way.
Website Ownership Verification
If you want to be sure that your publishers actually own the websites they claim to own, you should enable the "Website ownership verification" option.
This option will force publishers to verify their websites. After submitting a new website or changing the URL of the existing website, publishers will be redirected to the verification page.
There are three verification options available (commonly accepted by the industry):
- HTML-file upload
Publishers need to upload a special file to the root directory of a website.
- HTML META-tag
Publishers have to insert META-tag into the HTML code of the main page of a website.
Publishers have to add a TXT record to the DNS zone of a domain.
After successful verification, the website status will be switched to either approved or pending, depending on your "Websites post-moderation" settings.
Domain Matching Check
If you want to be sure that the ad code is placed on the correct website and that the traffic actually comes from this website, you should also enable the "Domain matching check" option.
Please note that changes in these settings do not apply to existing zones and will only affect new ones. If you'd like to make changes to existing zones, please do it manually via Anti-fraud options or by using bulk zone updater (Services / Zone update menu).
Due to technical limitations, this check only works correctly with JS-based ad code (e.g., banners, wrappers). Unfortunately, it may not work correctly with other ad formats (e.g., direct links, VAST).
We do not recommend using this option on zones used as RTB/XML endpoints because it may cause traffic loss.
Be aware that changes in anti-fraud settings do not apply to existing zones automatically and will only affect new ones.
If you'd like to make changes to existing zones, please do it manually via Anti-fraud options or a bulk zone updater (Services / Zone updater menu).