Online dating has come a long way from its somewhat dubious beginnings — it’s likely a friend, relative, or even yourself has found a partner through a matchmaking site. Much of that can be attributed to online dating services becoming increasingly effective at identifying matches, a feat that requires a sophisticated understanding of the psychology of attraction, interface design, and of course, a wealth of user data. In recent years, services such as OKCupid have touted their data-driven results, claiming to have developed algorithms that crack the mysteries of love and desire. But in a recent interview with GigaOm, Zoosk co-founder Shayan Zadeh states that such claims are overplayed.

Big data plays an important role in how Zoosk matches mates, Zadeh acknowledges, but it’s difficult — if not impossible — to engineer serendipity. “The magic is humans talking to each other…not computers picking [who you'll date],” he says in the interview. To that end, Zoosk emphasizes data-driven user experience design, to build a service that not only performs matchmaking services, but also facilitates chance encounters and connections among users. This requires a user experience that feels less like speed dating, and more like a comfortable nightclub where people want to hang out in and chat. To find out more about how Zoosk uses its data to encourage serendipitous meetings between folks looking for love, check out the interview at GigaOm.