Affiliate marketing is stagnant and ripe for innovation. In-content “affiliate” or “performance” marketing -- driving traffic through product links embedded in original content -- is one of the oldest business models on the web and it has barely changed in a decade. This year, 2014, change is emerging.
A new affiliate exchange, bringing merchants and publishers together to buy and sell clicks in real time, launched in the summer of 2013. This new content-driven commerce exchange is picking up steam. As a result of this exchange and others like it, publishers will unlock billions of dollars in revenues in the next five years. Merchants will be able to reliably and predictably buy product specific content-driven commerce clicks on demand and at scale.
Affiliate marketing is plagued by a reputation for fraud. It is also dominated by coupon sites that often do little more for merchants than create margin pressure. Equally problematic, affiliate marketing suffers from mind boggling fragmentation and complexity. Tens of thousands of merchant programs are spread across dozens of affiliate networks in the US alone. Much of the work required of publishers for earning from clicks they drive to merchants remains manual and error prone. Until recently, real- time bidding for contextually-relevant product placements within original content hasn’t been possible.
Sites like the New York Times don’t monetize through affiliate marketing not out of high-minded editorial integrity but because old school affiliate marketing isn’t worth the trouble. For these reasons, affiliate marketing has never achieved the economies of scale of either search or display advertising.
In 2014, this has all started to change. A combination of big data, Natural Language Processing, and powerful predictive analytics has automated away the complexity. These technologies sound complex. In fact, they simplify all of the messy pieces that comprise creating, pricing, and filling affiliate inventory in a rational two-sided marketplace. Both the buyers (online merchants) and the sellers (online publishers) of affiliate clicks are benefiting.
This has allowed for the emergence of the first ever content-driven commerce exchange. In this exchange publishers auction clicks on product links embedded in their content to the highest bidder that sells the product (A Nikon D5300 camera is the same camera if you buy it at BestBuy or on Amazon). By bidding to buy these in-content shopping clicks, merchants are winning more sales. At scale, this shift will boost publishers’ commerce-based revenues by double-digit to triple-digit percentages. Online merchants finally gain reliable, predictable access to commerce driven by trusted content and the aggregated audience of in market shoppers.
The drivers for this change are self-evident to every online publisher. Today, publishers sell clicks on product mentions embedded in their original content with no idea how much the average click yields in revenue. They cannot keep track of changes in commission structures across dozens of affiliate networks or direct performance marketing relationships with merchants. They cannot predict in advance whether traffic to one online merchant will convert at a higher rate than traffic to another.
Publishers should not be expected to build sophisticated models to predict which merchant will pay the most for, let’s say, a German visitor on a mobile phone clicking on a deep link to a pair of dress shoes. The results are woeful inefficiency. A publisher trying to manage affiliate marketing manually is lucky to monetize a third of their commerce clicks, and at rates that drastically undervalue their worth. This is a classic yield management problem, long ago solved for both search and display.
Merchants also suffer. There is no unified marketplace, no NASDAQ or DoubleClick or AdWords for clicks from content. Today, merchants must navigate the existing universe of sophisticated click traffickers. These include the classic affiliate networks, comparison shopping engines and countless other niche players. Successful integration into an affiliate network is neither easy nor fast. As a result, switching costs are high. In the end, merchants that need to buy extra shopping clicks struggle to find them.
The solution for affiliate clicks is a platform where big data trains ever-smarter models that drive advertising automation. What humans see as chaos -- a pool of content and clicks fragmented across a giant mass of affiliate networks -- computers see as data -- normalized, structured, relational data.
This allows software to create linkages between content sites and commerce sites and to price those linkages efficiently. This efficiency will place the best merchant offers on clicks delivered by the publishers with the best audience. Today, the best models can automatically identify product mentions with high precision. Predictive pricing models select the most economically rational link based on factors such as commission fee structure and merchant conversion rates. Software automatically embeds that link, routing traffic to the most competitive retailer.
The first content-driven commerce exchange opened for business in June. Every day, it auctions off thousands of clicks on product mentions to eager merchants. Publishers see earnings per click that were 200% to 300% higher, on average, as compared to clicks not flowing through the exchange. A number of merchants eagerly jumped in to gain access to one of the largest aggregated pools of content-driven shoppers on the Web today. Most importantly, the whole process was orderly and painless on both sides. It’s the future of content-driven commerce and it’s inevitable.