htened during the October 2003 upgrade to the hoplink system. Although these measures do not constitute perfect solutions to the ongoing problem of link theft, the progressive enhancement of the referral system is helping to deter the casual hijacker.

Merchants can also play a role in protecting their affiliates against referral losses, both through education and through the use of protective technologies.

A small minority of merchants employs the somewhat drastic step of screening every purchase and validating the referring affiliate (if any) prior to the delivery of the product. The validation occurs in real-time, using a database of registered affiliates. If the referral comes from a known affiliate, the product is delivered in the normal manner. But if the referrer is unknown - as would be the case when a link theft occurs - the buyer finds herself in the embarrassing position of explaining how the referral occurred. Unfortunately, the technical challenges in implementing and managing such a system with ClickBank are likely to exceed the benefits it would deliver

If you have visited any of the affiliate-related websites or read any of the various ebooks about affiliate marketing, you’ve almost certainly seen references to the thorny subject of link theft. Indeed, so contentious is this issue that it is frequently discussed in forums and a whole arsenal of third party weapons has evolved to help defeat it. But is the problem really as serious it is made out to be?

As an affiliate, it is almost impossible to quantify the amount of commission you lose through deliberate link hijacking or other, less malicious, forms of commission losses. You can accurately measure the number of referrals you make to any particular merchant or product and your sales stats will tell you what percentage of these referrals result in sales. But you have no way of knowing how many sales occur in a manner that somehow deprives you of your commission.

Even merchants, who see the other side of the business relationship, can only guess at the true extent of commission losses. When a merchant reviews his overall sales stats he will see two types of sales: those on which an affiliate commission was paid and those for which there is no known affiliate. Within the former group, the majority will derive from genuine affiliate referrals, but a percentage will be the result of link theft. Similarly, within the latter group, a proportion will derive from the merchant's own promotional efforts and the remainder from other forms of losses, including bypassing. But that's as far as the analysis goes. There is no accurate way of isolating link theft from genuine referrals and, likewise, no way to determine the level of other losses disguised as direct sales. Even a merchant who undertakes no active promotion of his own cannot be sure that all of his un-attributed sales arise out of referral losses. Some of them may come from unsolicited search engine listings or inbound links from other non-affiliated websites.

However, an interesting experiment was conducted by Bogdan Ravaru, author of The HTML Security Report, in which he created the conditions necessary to accurately measure link theft. After launching a new software product, using a newly established ClickBank account, he signed himself up as the sole affiliate. By not publicizing his affiliate program, he could be certain that there would be no other legitimate affiliates.

He then used paid advertising to generate a small but rapid influx of web traffic to his sales page. He was pleased with the sales results - 13 immediate buyers for his product - but was astonished to find that 46% of them were referrals from affiliates other than him. Clearly, 6 of the 13 buyers were already ClickBank affiliates who had used their own affiliate accounts to secure illicit discounts on their purchases.

Of course, this is just one isolated experience. But other marketers have conducted their own trials and tend to agree that the overall rate of commission losses may be as high as 35%. Clearly, the problem is serious enough to have a significant impact on affiliate income.

There is a consensus of opinion among marketers that link theft is worse on products aimed at the online marketing community - the suggestion being that marketers are knowledgeable about the affiliate process and thus better able to manipulate it to their advantage. There are also large variations in theft levels between different types of affiliate programs.

Affiliate networks, such as ClickBank, are much more susceptible to hijacking than standalone programs. This is because of the greater statistical likelihood that any given referred prospect will already be a program member. Few prospects would go to the effort of joining a new affiliate program, merely to secure an illicit discount, but many would be tempted by easy savings from an existing program membership.

Before ClickBank introduced its current hoplink procedure, it was possible for a link thief to misappropriate the commission on a purchase, simply by re-invoking the hoplink process using a manually entered URL. The ease with which a dishonest affiliate could "type and steal" made it an attractive target for casual thieves.

But this loophole was eliminated as part of a range of security enhancements introduced by ClickBank in August 2002. And referral security was further tightened during the October 2003 upgrade to the hoplink system. Although these measures do not constitute perfect solutions to the ongoing problem of link theft, the progressive enhancement of the referral system is helping to deter the casual hijacker.

Merchants can also play a role in protecting their affiliates against referral losses, both through education and through the use of protective technologies.

A small minority of merchants employs the somewhat drastic step of screening every purchase and validating the referring affiliate (if any) prior to the delivery of the product. The validation occurs in real-time, using a database of registered affiliates. If the referral comes from a known affiliate, the product is delivered in the normal manner. But if the referrer is unknown - as would be the case when a link theft occurs - the buyer finds herself in the embarrassing position of explaining how the referral occurred. Unfortunately, the technical challenges in implementing and managing such a system with ClickBank are likely to exceed the benefits it would deliver

those on which an affiliate commission was paid and those for which there is no known affiliate. Within the former group, the majority will derive from genuine affiliate referrals, but a percentage will be the result of link theft. Similarly, within the latter group, a proportion will derive from the merchant's own promotional efforts and the remainder from other forms of losses, including bypassing. But that's as far as the analysis goes. There is no accurate way of isolating link theft from genuine referrals and, likewise, no way to determine the level of other losses disguised as direct sales. Even a merchant who undertakes no active promotion of his own cannot be sure that all of his un-attributed sales arise out of referral losses. Some of them may come from unsolicited search engine listings or inbound links from other non-affiliated websites.

However, an interesting experiment was conducted by Bogdan Ravaru, author of The HTML Security Report, in which he created the conditions necessary to accurately measure link theft. After launching a new software product, using a newly established ClickBank account, he signed himself up as the sole affiliate. By not publicizing his affiliate program, he could be certain that there would be no other legitimate affiliates.

He then used paid advertising to generate a small but rapid influx of web traffic to his sales page. He was pleased with the sales results - 13 immediate buyers for his product - but was astonished to find that 46% of them were referrals from affiliates other than him. Clearly, 6 of the 13 buyers were already ClickBank affiliates who had used their own affiliate accounts to secure illicit discounts on their purchases.

Of course, this is just one isolated experience. But other marketers have conducted their own trials and tend to agree that the overall rate of commission losses may be as high as 35%. Clearly, the problem is serious enough to have a significant impact on affiliate income.

There is a consensus of opinion among marketers that link theft is worse on products aimed at the online marketing community - the suggestion being that marketers are knowledgeable about the affiliate process and thus better able to manipulate it to their advantage. There are also large variations in theft levels between different types of affiliate programs.

Affiliate networks, such as ClickBank, are much more susceptible to hijacking than standalone programs. This is because of the greater statistical likelihood that any given referred prospect will already be a program member. Few prospects would go to the effort of joining a new affiliate program, merely to secure an illicit discount, but many would be tempted by easy savings from an existing program membership.

Before ClickBank introduced its current hoplink procedure, it was possible for a link thief to misappropriate the commission on a purchase, simply by re-invoking the hoplink process using a manually entered URL. The ease with which a dishonest affiliate could "type and steal" made it an attractive target for casual thieves.

But this loophole was eliminated as part of a range of security enhancements introduced by ClickBank in August 2002. And referral security was further tightened during the October 2003 upgrade to the hoplink system. Although these measures do not constitute perfect solutions to the ongoing problem of link theft, the progressive enhancement of the referral system is helping to deter the casual hijacker.

Merchants can also play a role in protecting their affiliates against referral losses, both through education and through the use of protective technologies.

A small minority of merchants employs the somewhat drastic step of screening every purchase and validating the referring affiliate (if any) prior to the delivery of the product. The validation occurs in real-time, using a database of registered affiliates. If the referral comes from a known affiliate, the product is delivered in the normal manner. But if the referrer is unknown - as would be the case when a link theft occurs - the buyer finds herself in the embarrassing position of explaining how the referral occurred. Unfortunately, the technical challenges in implementing and managing such a system with ClickBank are likely to exceed the benefits it would deliver

ng a newly established ClickBank account, he signed himself up as the sole affiliate. By not publicizing his affiliate program, he could be certain that there would be no other legitimate affiliates.

He then used paid advertising to generate a small but rapid influx of web traffic to his sales page. He was pleased with the sales results - 13 immediate buyers for his product - but was astonished to find that 46% of them were referrals from affiliates other than him. Clearly, 6 of the 13 buyers were already ClickBank affiliates who had used their own affiliate accounts to secure illicit discounts on their purchases.

Of course, this is just one isolated experience. But other marketers have conducted their own trials and tend to agree that the overall rate of commission losses may be as high as 35%. Clearly, the problem is serious enough to have a significant impact on affiliate income.

There is a consensus of opinion among marketers that link theft is worse on products aimed at the online marketing community - the suggestion being that marketers are knowledgeable about the affiliate process and thus better able to manipulate it to their advantage. There are also large variations in theft levels between different types of affiliate programs.

Affiliate networks, such as ClickBank, are much more susceptible to hijacking than standalone programs. This is because of the greater statistical likelihood that any given referred prospect will already be a program member. Few prospects would go to the effort of joining a new affiliate program, merely to secure an illicit discount, but many would be tempted by easy savings from an existing program membership.

Before ClickBank introduced its current hoplink procedure, it was possible for a link thief to misappropriate the commission on a purchase, simply by re-invoking the hoplink process using a manually entered URL. The ease with which a dishonest affiliate could "type and steal" made it an attractive target for casual thieves.

But this loophole was eliminated as part of a range of security enhancements introduced by ClickBank in August 2002. And referral security was further tightened during the October 2003 upgrade to the hoplink system. Although these measures do not constitute perfect solutions to the ongoing problem of link theft, the progressive enhancement of the referral system is helping to deter the casual hijacker.

Merchants can also play a role in protecting their affiliates against referral losses, both through education and through the use of protective technologies.

A small minority of merchants employs the somewhat drastic step of screening every purchase and validating the referring affiliate (if any) prior to the delivery of the product. The validation occurs in real-time, using a database of registered affiliates. If the referral comes from a known affiliate, the product is delivered in the normal manner. But if the referrer is unknown - as would be the case when a link theft occurs - the buyer finds herself in the embarrassing position of explaining how the referral occurred. Unfortunately, the technical challenges in implementing and managing such a system with ClickBank are likely to exceed the benefits it would deliver

about the affiliate process and thus better able to manipulate it to their advantage. There are also large variations in theft levels between different types of affiliate programs.

Affiliate networks, such as ClickBank, are much more susceptible to hijacking than standalone programs. This is because of the greater statistical likelihood that any given referred prospect will already be a program member. Few prospects would go to the effort of joining a new affiliate program, merely to secure an illicit discount, but many would be tempted by easy savings from an existing program membership.

Before ClickBank introduced its current hoplink procedure, it was possible for a link thief to misappropriate the commission on a purchase, simply by re-invoking the hoplink process using a manually entered URL. The ease with which a dishonest affiliate could "type and steal" made it an attractive target for casual thieves.

But this loophole was eliminated as part of a range of security enhancements introduced by ClickBank in August 2002. And referral security was further tightened during the October 2003 upgrade to the hoplink system. Although these measures do not constitute perfect solutions to the ongoing problem of link theft, the progressive enhancement of the referral system is helping to deter the casual hijacker.

Merchants can also play a role in protecting their affiliates against referral losses, both through education and through the use of protective technologies.

A small minority of merchants employs the somewhat drastic step of screening every purchase and validating the referring affiliate (if any) prior to the delivery of the product. The validation occurs in real-time, using a database of registered affiliates. If the referral comes from a known affiliate, the product is delivered in the normal manner. But if the referrer is unknown - as would be the case when a link theft occurs - the buyer finds herself in the embarrassing position of explaining how the referral occurred. Unfortunately, the technical challenges in implementing and managing such a system with ClickBank are likely to exceed the benefits it would deliver

htened during the October 2003 upgrade to the hoplink system. Although these measures do not constitute perfect solutions to the ongoing problem of link theft, the progressive enhancement of the referral system is helping to deter the casual hijacker.

Merchants can also play a role in protecting their affiliates against referral losses, both through education and through the use of protective technologies.

A small minority of merchants employs the somewhat drastic step of screening every purchase and validating the referring affiliate (if any) prior to the delivery of the product. The validation occurs in real-time, using a database of registered affiliates. If the referral comes from a known affiliate, the product is delivered in the normal manner. But if the referrer is unknown - as would be the case when a link theft occurs - the buyer finds herself in the embarrassing position of explaining how the referral occurred. Unfortunately, the technical challenges in implementing and managing such a system with ClickBank are likely to exceed the benefits it would deliver.

If all else fails, the affiliate can take her own steps to protect and survive. The simplest of these involves only minor changes to the HTML code used in the web pages containing referral links and other techniques, including the ever-popular link cloaking, are in widespread usage.

But, despite the considerable selection of protective technologies employed by ClickBank, its merchants and their affiliates, none is foolproof. For example, a determined link fraudster can defeat every known defensive system simply by deleting her ClickBank cookie file prior to making a purchase. If an affiliate is sufficiently savvy and committed to gaining an undeserved commission, nothing will stand in her way. Therefore, as with any business problem, we must avoid the temptation to become obsessed with referral security at the expense of our other profit-making activities.