clickfraud

This page considers clickfraud, in particular bogus clicks on online advertisements.

It covers -

• introduction - what is click fraud?
• three forms - varieties of clickfraud
• prevalence - how much click fraud and where is it coming from
• regulation - responses to clickfraud

     introduction

As preceding pages of this profile have indicated, advertising is a significant part of the 'internet economy'. Businesses, government agencies and other entities use a range of online mechanisms to secure the attention of the online population, including -

• static and dynamic banner advertisements (eg animations and even videos)
• paid entries in directories
• paid placement of entries in or adjacent to lists of search results displayed by search engines
• paid placement of links in blogs

Some of those mechanisms simply inform/entertain the viewer. Others act as gateways to independent sites (in particular the site maintained by the advertiser) or services. They may be aimed at an undifferentiated mass audience or instead narrowly targeted at a specific demographic.

Advertisers typically pay on the basis of

• overall visits to a site (or to a particular part of a site), using the circulation or exposure model established by newspaper/magazine publishers and broadcasters - how many people were likely to see the publication, hear the radio broadcast or watch the television program
• indicators that the audience has sighted the specific advertisement/entry - demonstrated by the audience clicking the particular ad, directory entry or other hyperlink.

Advertising may be arranged directly or through an agent, with traditional media buying services and online specialists for example representing individual advertisers in 'buying space' online and acting as intermediaries between site owners and advertisers in making payments and ensuring compliance with contracts.

Placement costs - how much the marketer pays for an appearance on someone else's online real estate - vary from fractions of a cent per eyeball (or per click) to several hundred dollars per click or eyeball.

Clickfraud - sometimes genteelly characterised as 'invalid clicks' - subverts those mechanisms. It encompasses illicit clicks by site owners (seeking revenue to which they are not entitled) and by competing advertisers (seeking to increase an opponent's costs). It also encompasses mundane fraud where an advertiser or agent denies a site owner legitimate revenue by claiming that clicks have not occurred.

     three forms

Clickfraud takes three basic forms, with fraud by -

• site owners
• competitors
• advertisers or their agents

Click fraud by site owners involves that owner recurrently 'clicking' on an advertisement or paid link that appears on the site, with the aim of deceiving the advertiser by significantly increasing the number of clicks and thence boosting the payment due to the owner for use of the space.

The actual clicking typically is not done by the owner but instead is outsourced to local itinerants (for example students or housewives are paid to spend a day clicking) or people offshore, with some industry figures accordingly joking about Indian 'clickshops' - the clickfraud counterpart of call centres. It may also be undertaken by software.

Fraud by competitors (and by hacktivists or merely by people acting out of malice rather than a commercial interest) malicious reflects the fact that the advertiser pays for the advertisement or preferential placement in/around a set of search results.

It has been claimed that some people attempt to burn a competitor's budget, either on the basis that with enough clicks the advertiser will run out of money (and thereupon be relegated to a lower placement) or will simply become disillusioned and withdraw after a large number of clicks is not reflected in increased sales or queries by potential customers/supporters.

Such a claim is problematical and, if true, will presumably affect SMEs and small NGOs rather than major corporations. We don't suggest trying to click Toyota or Macquarie Bank into receivership.

Fraud by advertisers or their agents is far simpler, based on the naivety

     prevalence

What is the scale and scope of clickfraud? Is it increasing? Is it seriously inhibiting electronic commerce? Is it worse in some jurisdictions than others? Is it predominantly affecting particular types of advertising or categories of sites, for example those of SMEs?

The answers to those questions are contentious.

Disagreement reflects the absence of authoritative academic studies and independent commercial auditing. There are few formal standards and the immaturity of the online advertising and metrics industries (or merely their self-interest, as there are commercial incentives to maintain a 'digital wild wild west') has not resulted in the counterparts of the newspaper/magazine circulation and broadcast audience reports found in the offline environment.

Online enthusiasts note that problems are evident offline, with recurrent disputes about

• whether people actually sight content in particular publications (as distinct from buying the newspaper)
• whether people are in the same room and concentrating when a broadcast takes place, something of increasing concern with uptake of PVRs
• the authority of competing commercial ratings agencies and devices such as peoplemeters

It has been variously claimed that bogus clicks amount to 10%, 30% or even "much greater than 50%" of all advertising clicks or clicks for paid placement in search results. A definitive answer is unavailable.

It is clear that clickfraud is occurring in Australia. However the extent of that fraud and its impact is unclear.

Does clickfraud make commercial sense? One response is that perpetrators might get a greater return from making money (or voicing a protest) the old fashioned way.

     regulation

Is clickfraud unethical but legally permissible or merely beyond effective regulation?

For some cyberlibertarians (or merely people whose vision of cyberspace does not encompass commercialisation of the net) clickfraud is not an issue. Some business figures have acknowledged that it exists but dismissed its importance, asserting that it does not occur on a scale and with a frequence to merit serious attention.

As of 2006 no nation appears to have enacted a statute specific to clickfraud. One reason is presumably that existing common and statute law is considered to provide appropriate coverage. In principle clickfraud clearly is an offence, given that it improperly deprives an entity of revenue and may involve misrepresentation.

In practice major sites and search engines such as Google are using traffic analysis technologies to identify fraud. Such tools will for example flag unusual peaks in visits to a page or ad (suggesting that a bot or individual has settled down for a quota of 2,000 clicks), identify that most visitations are coming from a particular IP address or indicate that visitations are too uniform to be real (for example that a bot has been programmed to hit a particular page or link eery 1.5 seconds for two hours).

    next page  (placement)