Earlier this year, I personally led the charge on implementing impression based monitoring technology across Gourmet Ads. The technology now measures every single impression served both across our directly managed network of food sites and when we buy impressions in the Appnexus Exchange. The main goal of this monitoring was to ensure our client’s advertising was being viewed by real humans and not computerized bots.
The move towards impression monitoring was triggered after we met with some of our best clients ranging from direct customers to advertising agencies and programmatic buyers. During these meetings, we asked what their greatest issue is right now. The feedback we received in these meetings was invaluable. Consistently, our clients Number 1 issue they face across the industry is the minimization of buying fraudulent sites and non-human traffic for their campaigns. So to ensure that Gourmet Ads was compliant, we implemented a solution in early April.
Publishers doing the Wrong Thing are not Welcome @ Gourmet Ads
Since then, we’ve taken a Zero Tolerance to sites we directly manage showing suspicious behaviour. We’ve to-date expelled 10 sites now from our directly managed network and will continue to do so when we see suspicious behaviour from a publisher. Not surprisingly, after removing these suspicious sites we saw our average click through rate drop across the network. But in reality, these were never “real” clicks to begin with. We are tracking unusual click patterns, impression bursts and spikes, Ad Tags run over multiple URLs, IP tracking and many other factors that could indicate unusual behaviour.
Publishers that insist on utilizing of non-human traffic and bots are not welcome at Gourmet Ads. What makes me angry is fraudulent sites are effectively stealing revenue from legitimate publishers who work hard every day to create great content. Apart from impression based monitoring, we now have three measures in place to do our best to prevent fraud sites from ever joining Gourmet Ads, however it’s not perfect and I’m sure one or two will still slip through. Rest assured if something is brought to our attention as it has slipped through the gaps, we’ll dig into the data, learn how it was missed and take the necessary action.
A Solution for Direct Managed Sites & Exchange Bought Impressions
On the audience extension side, our operations team continues to build a significant black list of sites that we will never buy inventory on when doing retargeting for clients. Feedback we had early on from fraud experts in the advertising space said that food sites (as well as lifestyle and gossip sites) are consistently favoured by fraudsters, thus putting Gourmet Ads right in the sights. Recipe sites in particular are relatively easy to create by buying a domain name, then downloading a database of recipes and photos and putting them in an open source CMS.
Since we embarked on this process we have discovered hundreds of fraudulent food sites and few rather large scale “clusters” of fraudulent recipe sites. Some simply duplicate content and change the domain name and logo, whilst others are sophisticated groups of sites that interlink between each other. We’ve also discovered one food site that looks like legitimate site on face value, but behind the scenes had pages running over 100 ad tags on a single page. When relevant, we also proactively encourage our ad serving partners to block and/or reject sites where we have solid evidence of fraudulent behaviour. Given our access to impression level data and the ability to cross reference it from other sources, these sort of activity stick out rather quickly.
To date our clients have been extremely happy with the results. Not only that, our clients know when they are buying Gourmet Ads they are buying quality and not something fraudulent.
I‘d like to think that Gourmet Ads is proactively leading the charge on combating fraud in the ad network space. I hope that other ad networks and major exchanges also put their foot down and take a similar stance on fraudulent behaviour with the view of eradicating it across the industry. As I said recently while presenting at Ad Week in NYC, I believe this is a massive and ongoing issue, somewhat similar in nature to credit fraud in the ecommerce space and we all need to actively do what we can to remove it.
If you’re an advertiser or programmatic buyer and would like to learn more about our solution we have implemented, please contact us and we’d be happy to walk you through it.