At the same time, latinos were generally favored by both white men and women willing to out-date. In 2014, kang zhao at the university of iowa constructed a new approach based on the algorithms used by amazon and netflix, based on recommendations rather than the autobiographical notes of match seekers. Users' activities reflect their tastes and attractiveness, or the lack thereof, they reasoned. This algorithm increases the chances of a response by 40%, the researchers found. E-commerce firms also employ this "Collaborative filtering" technique.

In contrast, white men are excluded by 76% of black women, 33% latinas, and only 11% asian women. Similarly, 92% of white women exclude black men, 77% exclude latinos, and 93% exclude asian men. 71% of black men, 31% of latinos, and 36% of asian men excluded white women. In short, after opposite-sex members of their own group, white men were open to dating asian women, and white women black men than members of other racial or ethnic groups.

Some sites are completely free and depend on advertising for revenue. Others utilize the freemium revenue model, offering free registration and use, with optional, paid, premium services. Still others rely solely on paid membership subscriptions.

(see comparison of Speed Dating services.) some have a broad membership base of diverse users looking for many different types of relationships. Other sites target highly specific demographics based on features like shared interests, location, religion, sexual orientation or relationship type. Online dating services also differ widely in their revenue streams.

Nevertheless, it is still not known what the algorithm for finding the perfect match would be. A great diversity of online dating services currently exists.