Personalization Makes the Customer Grow Fonder
Singapore-based dating app Paktor uses profile and behavioral data combined with testing to guide its personalization efforts.
Brand experiences and customer preferences aren't always a match, but marketers can use personalization to prevent customers from swiping left. In fact, marketers for Singapore-based dating app Paktor did just that by optimizing the customer experience through testing—and saw the additional benefit of boosting subscription rates as a result.
A love affair with data
Paktor may be in the business of playing Cupid, but it sustains its mobile matchmaking service by acquiring customer data that allows its marketers to use extensive personalization in its marketing.
One way Paktor obtains this information is through its users' profiles. After customers download the free app, they're asked to give Paktor access to their location data, which enables the app to show them profiles of potential matches within their area. Then, they're prompted to log in via Facebook. Once they do, their Paktor profile is automatically updated with their public Facebook profile information and photo. But users can update this information or add details about themselves such as height, occupation, and education level, says Anmol Mohan, Paktor's head of data analytics. “A profile is really important for us,” he says.
Users can also specify what kind of partner they're looking for by indicating within their settings their preferred gender, age range, country of residence, height, job, education, and language.
In addition to collecting profile data, Paktor tracks users' behavioral data as they use the app. Once users create their profile, they can look at the profiles of potential matches who meet their criteria. If someone isn't their type, they swipe left; if they're interested in getting to know someone better, they swipe right. If a potential match also swipes right on their profile, they're considered a match and added to both parties' match list. They can also begin messaging each other.
Paktor tracks these activities. Marketers for the dating app can then use this data to segment its users and send personalized messages to encourage them to use the app. For example, if Paktor detects that a user is dormant but that he has a match, it will send that user an email inviting him to revisit the app to chat with his potential match, Mohan explains.
A new relationship with monetization
Paktor is a free dating app, but it is a business, so it looks for ways to monetize the app. Last year the company launched a points-based monetization offering.
Users can download the app, view profiles, and message matches for free. They can also use Paktor points to send a direct message request to people who didn't respond to their match request or didn't consider them a match. Plus, they can use these points to send gifts. Users can purchase points a la carte, or can buy a one-, three-, or 12-month subscription, which grants them 1,000 points daily and provides them access to additional features, such as the ability to view suggested profiles or have access to premium filters.
But few users were discovering the benefits of subscriptions on their own, Mohan says. To drive subscriptions, Paktor's marketers focused on personalization. They worked with mobile app optimization tool provider Apptimize to conduct an A/B test for a pop-up message. The system presented a message to users who viewed 10 or more potential matches; the copy was tailored depending on whether they had been highly engaged and swiping right or they had been disliking several profiles and swiping left to reflect their moods. Either way, the pop-up directed users to the subscription page.
“It's important to understand the behaviors of the users on the app," Mohan says. "Our app is in a constant state of experimentation.”
A long-term commitment to testing
After conducting the test, Paktor experienced a 10.35% increase in subscriptions. And even though its a la carte purchases decreased, Paktor saw a 17% increase in average revenue per user and a 100-time increase in revenue in one year.
Mohan says Paktor has continued to experiment with its personalization capabilities. Marketers there test the frequency of how often to refresh the profiles users see; they've also tested the ability to share profiles. It's this testing that allows Paktor's marketers to understand user expectations and then use personalization to drive success.
“The key is personalization,” Mohan says. “You have to dig deep in your data to see what kind of feature is being accepted by which kind of segment.... You can keep users engaged and at the same time give them what they really want in the app.”
As for marketers who'd like to generate their own app success, Apptimize's CEO and Cofounder Nancy Hua advises them to focus on their existing customers—not just their new ones—and on creating the best user experience for them.
"You can't really spend money on acquisition right now without nailing retention and making sure that the customer is actually performing the action that they came into the app to do," she says. "Marketers have to figure out how to use the information that the user is giving them about what their intentions are so that they have a good product experience."