Zoosk set out to help users create richer connections at a faster pace. We noticed that some users were having trouble initiating conversations, and the users that did send messages had a lower than expected response rate. We challenged ourselves to develop a tool to solve this.

‍

Before deep-diving into this project, we wanted to confirm that the challenge we rose to was an issue that users faced. We released internal and external surveys to vet out what users were having trouble with the most via online dating apps, specifically our product. The information we received confirmed that a critical point of friction was messaging. We then began to dive into the Zoosk product and audit our messages. Many male users sent one-word or multi-paragraph messages, which yielded no response.

‍

‍

Both messaging strategies never reflected that the male user viewed the other person's profile. So I developed an experiment we later coined "Quick hello (QH)." Companies have attempted this before by offering a set of canned" hellos." However, approaching messaging this way potentially comes off as cold and robotic. The difference here is that we developed an AI system for user's profiles and, based on your inputted information, would provide message options that were unique and relevant.

We followed and tested this service over several months and what we found was extremely exciting. Not only were users engaging in the message options, but the feedback we received mentioned using these cues in real life! For example, users who usually found sending messaging and starting conversations difficult now use QH as a starting point. Even users who opted against using the feature still used our cues as inspiration for their own messages.