Earlier in 2016, one of our agency clients approached us to help them promote an event taking place in the UK. We had integrated Chrysal.is in order to better measure the performance of the campaign, and double check what was being provided by the networks for the campaign.
As the client was advertising a specific niche event, we were tasked with setting up and running advertising campaigns to drive traffic and ultimately ticket sales to the event. One of the five adnets we worked with was a relatively new specialist advertising which had a large number of publishers in the niche we were targeting.
So we set up the campaign, set up tracking paramaters to measure the traffic sources (unknown to us as is standard when dealing with many adnets) and used Chrysal.is to optimise it from day one. As the days progressed, we started to see some unusual patterns in the click data, and noted it for investigation.
Some of these traffic sources had pretty reasonable conversion rates, however as it was a cpc network with a pre-paid deposit, we were able to identify three distinct types of anomalies, flag these to the network, and get credits back on the campaign.
So what was happening? A variety of things. Firstly, there were the amateurs, people clicking their own ads on multiple devices across consecutive days, but outside of the normal 24 hours frequency capping that networks use before displaying ads. Some of these also had it programmed to do so.
In other cases, you had the same people clicking ads across several different sites they had on the same network, multiplying the budget drain.
In other cases things were a bit more nuanced. We’d see lots of different traffic coming in from a particular publisher, but through the behaviour of one or two users, we could figure out what was going on, in this case a pay to click website which was cycling people through ads, giving them a ten second cooldown, filling in a captcha, and repeating the cycle on another link.
But as all publisher id’s relate to a website, we could then challenge all of the rest of the traffic. After all if three people show you how everyone else is acting, and you have the data to prove it, adnets will prefer to give you the credit back, rather than you draw down your already deposited funds, or refuse to pay all or part of an invoice.
And our findings? Across four advertising campaigns in different languages for the same product, 6.7% of the publishers soaked up 39% of the advertising budget with invalid traffic, which was credited back. In addition to this, a further 20.8% of traffic was identified as being plain old fashioned waste, not doing anything. Post-correction, the remaining budget with the topup were focused on the traffic sources delivering registrations and sales for our client.
So if you’d like the kind of capability in your organisation to properly enforce your affiliate program terms or advertising insertion orders, get in touch to talk to us about Chrysal.is.