Identifying the Ad Network Responsible for a Rogue ad (porn, etc.)

Published by: Darren Kelly , Photobucket
Published on: July 11, 2011

We are looking to see if anyone out there has a sound methodology with best practices on identifying rogue ads by identifying the ad network and the specific ad tag they are coming through. What software are you using? What are the steps you take to identify and grab all the relevant data? Even if there are multiple 3rd party redirects, can you tie it back to the specific ad tag running in your ad server?

Comments

Darren, it's clear that there are a lot of options for trying to identify rogue ads and general poor behavior by 3rd party ad tags.
For manual testing people have mentioned and/or I've used:
- Firebug
- The Fiddler
- Charles
- HTTPSpy
This tools require a somewhat technical user to find the source, and require that you actually managed to reproduce the ad on your site.

There are a number of automated tools that have come out in the last year, and they take different approaches, so if your desire is to automated this it's worth talking to all of them, the main one's being (alphabetically)
Adometry TagScan (http://www.adometry.com/publishers/adtag-analytics.php)
The Media Trust (http://www.themediatrust.com/media-scanner.php)
Krux Digital (http://www.kruxdigital.com/)

There really are key differences in how these companies approach the problem in terms of functionality, coverage, implementation, etc.
I know that two of these vendors are attending AdMonsters in San Diego next week. I'll be there representing Adometry.

Totally forgot, that adxpose (now part of Cosmcore) has a pretty good service for identifying who served what and from where. You need to load code with each ad, but its pretty easy.

https://www.themediatrust.com/ is also well known for compliance and verification.

Hi Darren,

The only tool I have seen that allows you to see what provider is at the end of a particular daisy chain is Krux Digital's Data Sentry. It will also show you which organisations were responsible for ushering in the offending campaign, as well as allowing you to actively block the domain from which it is served.

It depends on what ad server you are using. If you are using DART (the legacy product) you can actually view the Ad ID in the url (if you can mouse over and look in the status bar...or use Chrome Developer Tool...or something like that). I can never remember which of the many numbers it is, but it's one of the first few. Just have your numbers for the lines and sizes and start looking through them...you'll find it.

If you aren't using DART...not sure what to do...I HATE this aspect (amongst others) of networks...makes it so hard to have control over YOUR OWN SITE!!!

Hi Darren,

[heads up, we're a vendor and specialize in solving exactly this type of challenge :)]

In answer to your question - yes, there is a sound methodology: our system intelligently scans all ads from ad networks to detect rogue ads as well as privacy violations, regardless of the ad's original source. Check out http://www.dasient.com/product-services/anti-malvertising-protection/.

I'd be happy to answer any questions you have..
Regards,
Ariana

Hi Darren,
I think the only alternative to a having user do the identification through plug-ins or sniffing tools like fiddler/firebug is that your ad-server can do impression-based URL reports and has a screen-shot spider feature like e.g. Mediamind and Adform. But such tools only work if the ads are not served in Iframes, in which case the ad-servers are inevitably blind, and no other options exist. I am not very familar with networks in the US, but that is probably very likely to be the case.

Hey Darren,

While Firebug or fiddler work, the challenge we have with such a large decentralized network is that people from all over the country send us screen shots of a rogue ad. Which is really only the starting point because we need to be able to replicate the ad on the page or site they saw it and use fiddler/firebug to trace it back.

This is not always possible because the ad could be geo targeted or a one time instance.
So we try and train folks to use the tools like firebug/fiddler and send us the source code as well from a page if they see the issue.

I wish there was an easier way to do this. I know many of the RTB networks have browser plugins that identify an ad and the source for troubleshooting, but of course that person who is seeing he issue right then and there needs to have the plugin. I wish the adservers like DFP had a plugin like this to provide to folks,I've asked about this many times but haven't seen anything come through.

Hi Darren.

You can use web debugging tool like HTTP Watch/firebug/fiddler to identify from where the ad is serving.

Thanks,
Venkatesh.B

Rocket Fuel