previously would have been likely to sign up for those products using the desktop. As a result. as the mobile migration has been rapidly underway. we have seen our overall conversion rates challenged in those businesses. However. we expect to see this trend reverse itself for two reasons. First, the migration to mobile will either slow rapidly or end as mobile devices obtain a more stable level of penetration within the population. Second. we expect to be able to make significant product improvements to our mobile products over the coming years. driving meaningful conversion increases. Our mobile products are relatively earty In their development stages, as we only began to devote meaningful resources to optimizing these products In the last two years; in contrast, we have focused on optimizing our desktop products for increased conversion for many years. For example, during the period between January 1. 2008 and December 31, 2013, we Increased conversion on our Match product In the United States by more than 50% (when comparing desktop users coming in through direct domain channels, which we believe is the purest way to isolate the relationship between product changes and conversion improvements): those improvements came after more than 10 years of that product's existence (as opposed to the relatively short history of our mobile products). Therefore, based on our prior experience with product improvement, and the finite nature of the mobile migration, we believe the conversion challenges we have been facing as a result of the rapid mobile migration in these businesses will level off and then reverse. 64 Table of Contents Lower cost users. All of our brands rely on word-of-mouth, or free, customer acquisition to varying degrees. Word-of-mouth acquisition is typically a function of scale (with larger communities driving greater numbers of referrals). youthfulness (with the viral effect being more pronounced in younger populations due, in part, to a s