March 15, 2006 | Wednesday

Conversion data the real key to identifying click fraud

By Jon Myers, Director Of Search in News |Online Sales |Search Engines |Google |Search Expertise |Click Fraud

Kevin Newcomb writes at ClickZ that Google’s click fraud settlement:

can be seen as a victory for marketers in that Google’s offering a chance for some marketers to be reimbursed for invalid clicks over the last four years they’d otherwise not be eligible for. But the bigger winner here would appear to be Google itself. For just $90 million, a fraction of Google’s nearly $1.5 billion in net income from 2005, Google will essentially rid itself of all click fraud-related issues of the past four years.

This shows a step forward in the marketplace, giving search engine marketers a potential way forward in the ongoing click fraud issues that we are all currently dealing with in the UK and worldwide.

But to put this into perspective, we are talking about $90 million out of an approximate turnover in the last four years of $10.3 billion. This will mean Google is refunding 0.8% of its turnover. Looking at this in smaller numbers, it would be like spending £50,000 per month and receiving a fraud rebate of £400. The industry talks about fraud being up to in the region of 20+% (£10k on £50k ad spend); as Kevin writes above, this really is not, in the scheme of things, a great deal for Google to lose.

What makes it more difficult is that the search engines decide what a ‘valid click’ is. This decision is undertaken without being able to see the final ROI or conversion of any search campaign. Going forward, it is up to search specialists like us to drive conversion learning and feedback into the likes of Google. The ‘valid click’ can only be confidently defined by the search engines and those who have the data of how the click has converted (or not converted, as the case may be).