How much are you burning on Click Fraud? Recent reports suggest that if you are using paid advertising in search engines, it could be a double-digit percentage in some competitive market categories such as gambling, travel and financial services, but much lower in others. So it's a question worth asking with some companies spending hundreds of thousands or even millions on paid listings in search engines.

In this review I will look at the different types of click fraud, who the potential perpetrators are, and how to assess whether it is happening to you using different measures and software tools. This article is not an introduction to Pay Per Click marketing. If you aren't familiar with the concepts and practice of Pay Per Click marketing, you may want to check this recent WNIM article: http://www.wnim.com/archive/issue3705/emarketing.htm.

Click fraud on Paid Search advertising is nothing new - it has always existed. When I describe the concept of Pay Per Click marketing to any marketer new to it, the light-bulb soon comes on: 'so I can get someone to click on my competitor ads and bankrupt them?' they ask!

The growth in click fraud was brought home to me recently in a session by Mike Grehan of Smart Interactive at the recent E-metrics Summit (summarised in the Net Imperative piece referenced at the end of this article).

Mike revealed that large-scale click fraud is now rife in many sectors beyond the casino sector to which I thought it was largely limited. It is difficult to say exactly how big this problem is, but in some sectors, it is thought that click fraud can account for over 10% of clicks. With paid search becoming more competitive, this volume will certainly affect profitability. Not to mention the sheer waste of paying for ad spend with absolutely no return.

What is click fraud?
Mike explained there are two main types of click fraud:
" Click fraud - this is simply where a third-party generates clicks on a paid advertising link to drive up advertising costs with no intention to conduct business. These clicks can occur on ads placed on the search engine results page (search fake clicks) or ads placed on third party sites by a search ad network (content or contextual ad clicks)
" Impression fraud - generating false search queries

This type of fraud can potentially occur on any of the main ad networks such as Google Adwords (www.google.com/adwords), Overture (www.overture.com) or Miva (www.miva.com) which are used to place paid search ads.

Fake clicks or impressions can be generated either through manual clicks on the links or automated clicks through illicit software tools that generate clicks fake appearing to originate from different IP addresses, so that they cannot be detected by the search ad networks which watch for repeated clicks from the same source.

Another alternative is to pay a virtual team of clickers to click - this can be worthwhile with ads generating value of several dollars to site owners with context sensitive ads on third party sites.

This idea has been around for a while now - read the Times of India article - India's secret army of online ad 'clickers': http://timesofindia.indiatimes.com/articleshow/msid-654822,curpg-1.cms.

Know your enemy
So, who are the perpetrators of click fraud? They can include:
1. Direct competitors - this activity is typically manual and relatively small scale for ethical reasons, or perhaps more likely fear of legal recourse and reputational damage!

2. Affiliates - Affiliate marketing is the arrangement where owners of third-party sites gain commission for leads or sales gained on merchant sites by referring visitors to them. Affiliates tend to be sophisticated in search marketing and will, in some cases, use whatever it takes to compete with other affiliates or even with the brand sites which they promote. Affiliates are more likely to use automated click faking tools. Mike explained three common tactics for click fraud:

(a) Drive up advertising costs forcing you to increase your budget or drop out of competing on advertising

(b) Similarly, to deplete your budget and make the advertiser invisible for the rest of the day if they have set a maximum daily budget in Google Adwords, which is the typical practice

(c) Impression fraud - a more sophisticated approach which involves increasing ad impressions for rival advertisers, but not their clicks, so reducing their clickthrough rate. This can be done by the perpetrator suspending their own ads while impressions are generated. In Google, your ad position is dependent on clickthrough rates in addition to your bidded price per click. So, in Google, impression fraud may lead to your listing position falling and may even lead to dropping of ads by Google if it deems their clickthrough or relevance is too long.

3. Site owners - using Google Adsense or Overture Content Match, which display ads on third-party sites according to the context of content on your site (like my site, see www.davechaffey.com/Internet-Marketing) may manually or automatically click on links to generate revenue (unlike this site owner - no time, too ethical). This is sometimes known as Network Click Fraud. Incredibly around a half of Google's revenue is from these third-party contextual ads, yet those in the know sometimes find that clicks from these sources don't tend to convert at the same rate since customers typically aren't in research or buy-mode like they are within the main search engines - they are more impulse clicks. So look carefully at whether you are getting value from Google contextual Ad Sense ads or Overture Content Match ads. Warning, Google has recently won a judgement for $75,000 against a site owner who have used this approach on a large scale!

4. Customers - not strictly click fraud, but existing customers get into the habit of clicking on paid search ads to access a site but are not driving new business. Not much you can do about this, expect
perhaps changing the ad copy, educating about bookmarking or encourage clickthroughs via E-mail.

Checking for Click Fraud
So finally, what to do to avoid or at least minimise Click Fraud?

Well - first check your agencies or internal team's knowledge of click fraud - ask them about the different types of problem and see whether they are aware of them.

Second, ask them how they check for and how they will report to you on the extent of click fraud.

Mike Grehan suggested you should assess these metrics using a combination of your web analytics software and your software for managing your paid search campaigns:
" Average daily clicks (e.g. an increase from 10 per day to 100 per day suggests click fraud)
" Click volume trends by hour and day to see spikes in click volume
" Average page views per click (this will fall if people are not browsing around the site as would be expected from normal search behaviour)
" Average conversion rate per keyword, per click
" Click averages based on specific paid search positions
" Volume of clicks with zero conversions
" Off peak click times or dates
" Look for new competitors for your keyphrases and whether some have dropped out as clickthrough has increased
" Check the difference between clicks recorded by the ad network management system and actual clicks arriving at your site identifiable from your web analytics tool. If there are more actual clicks arriving than those reported by the ad network, they may have filtered some clicks out which they regard as fraudulent.

These measures can be used at an aggregate level which is straightforward or you should check whether you have the capability to report these for individual keyphrases or groups of keyphrases. For detailed analysis, it is necessary to have granular data about searches, so that you can report on searches performed for different keyphrases on the different ad networks. In some cases, it may be necessary to refer to raw log-files generated by the web server for each page downloaded to assess the frequency of clicks or even check for someone with a similar IP address or PC/browser setup (e.g browser, screen resolution, etc which are available from tag-based web analytics solutions).Your log files also become important when submitting your complaint to the search ad network, since they act as a form of proof of the click fraud. So for serious investigation of this problem, you probably need to combine different forms of web analytics tools such as tag or browser-based and server-based (See What's New in Marketing http://www.wnim.com/archive/issue04/emarketing.htm for more details on these tools).

To answer this second question, additional software tools specifically designed to tackle this problem may be available that can help. The best known is Who's Clicking Who, but Googling "Click fraud detection" reveals there are now several specific solutions available, plus more general web analytics tools.

What to do when you think you have been defrauded?
The search engine ad networks take click fraud seriously, since it undermines their proposition and can potentially do them reputational damage. This was highlighted at the end of last year, when Google Chief Financial Officer George Reyes said:

"I think something has to be done about this really, really quickly, because I think, potentially, it threatens our business model"

If you have good evidence to support your claim of click fraud you can typically expect a prompt response from the ad network. If you believe click fraud has occurred, provide a succinct summary of your evidence, but you will also need to back it up with reports from the ad networks reporting software and your web analytics software or log files. Grehan (2005) gives an example of a click fraud problem that was submitted to the network and a refund granted.

Click fraud - The future
Click fraud is another frustrating issue within E-marketing which seems as if it can only get worse, hence my highlighting it now. The potential for unscrupulous folks to make money through click fraud is just too high and it is relatively difficult to identify if the perpetrator is sophisticated in generating fake clicks across a range of IP addresses over a period of time. On the positive side, the search engines are taking legal action against some publishers generating fake clicks on contextual ads and some providers of automated click-faking tools. They are also likely to be boosting their click fraud detection tools, although they are highly unlikely to publicise this.

Some argue that click fraud is simply a cost that is borne by the advertiser and it should be factored into the overall campaign Return on Investment (ROI). If paid search is still delivering favourable ROI in comparison with other media, then what's the problem? The problem is that click fraud will always target particular phrases, which means that optimising on these phrases is difficult if they are not generating ROI. Perfectly valid phrases may be deleted as poor performers if click fraud is not taken into account.

Looking further forward, we may see a move to a form of paid search based on Cost per Acquisition (CPA) where the advertiser only pays where there is an event on the web page which demonstrates a lead or sale (for example, completing a registration or enquiry form). With the recent advent of Pay Per Call arrangements, this is effectively a CPA arrangement which is much less prone to fraud. Of course, the search ad networks will resist such a CPA arrangement since it will hit their revenues and it is only really appropriate for transactional e-retail. Maybe it could become an option though for certain e-retail-related phrases and advertisers would bid higher amounts if they are paying for a guaranteed sale.

Whatever happens, it will lead to some difficult strategic e-communications decisions. Perhaps increasing prevalence of click fraud suggests that companies should invest more to become more refined in their search engine optimisation which is not subject to this form of fraud. Another alternative is to increase reliance on affiliate marketing, since this is a CPA deal, and the risk of click fraud is borne by the publishers who are members of the affiliate network. One thing is for sure, it's a crazy situation where a billion dollar economy has grown up based around clicks on the humble hyperlink.

Next month's article
In the autumn issues of E-marketing Insights we will move away from the web for a while, looking at the latest developments in other new media such as mobile marketing, interactive Digital TV and radio.

References and Further reading
Grehan, M. (2005) The click, the fraud and the ugly side of search, Net Imperative, 23rd June 2005. www.netimperative.com.
http://www.netimperative.com/2005/02/01/click_fraud

Lee, K. (2005) Click Fraud - What it is, How to Fight It. . ClickZ : http://www.clickz.com/experts/search/strat/article.php/3483981

About the author
Dr Dave Chaffey is workshop leader for a range of one-day e-marketing training workshops from the The Chartered Institute of Marketing:
E-mail Marketing (www.cim.co.uk/0766)
Running Effective E-marketing Campaigns (www.cim.co.uk/0767)
Improving Your Results from Digital Marketing (www.cim.co.uk/1138)
Marketing Research Using the Internet (www.cim.co.uk/1135)

Go to www.cimtraining.com for course details and online booking.

Dave Chaffey is trainer and consultant for Marketing Insights Limited (www.marketing-insights.co.uk) and E-marketing Director at Ripe. He is a prolific e-business author whose books include 'Total E-mail Marketing', 'Internet marketing: Strategy, Implementation and Practice' and E-business and E-commerce Management.

Read Dave Chaffey's blog (www.davechaffey.com) for E-marketing Essentials - the 5 "must-read" articles about online marketing from the hundreds Dave reads each month.