Video Created by Eliana Feng @ Polaris Design.
Project Overview
Project Label Digital Product Design & Development
Clients In-house Product
Timeline September 2015 – Present.
Role & Responsibilities
Eliana Feng @ Polaris Design — Digital Product Design, Creative Direction, UX/UI Design, Visual Design, Front-End Development.
CodeBehind Development Team — Progressive Web Development.
Abstract
The key to solving the love equation, HotDate is a question game that focuses upon the critical aspects of social behavior and communication that drives human attachment. It is designed to be an extension and enhancement of current dating technologies. By reusing online matchmaking data — i.e OkCupid match questions, HotDate helps online daters continue the chemistry established by online trust proxies onto a real life situation, or more simply, the first date.
Domain Research
I started this project by conducting extensive research on the domain of online dating market. Based on my research analysis, I created a data visualization that maps out the landscape of 16 most popular online dating platforms currently live in the market (detailed below).
In order to dig into further the domain of online dating, I also read as many articles and book resources as I can. Two books that I read are highlighted below.
User Research
While trying to gain as much domain knowledge as possible, I also conducted extensive user research by interviewing online dating users and found out that most users face a similar problem:
A typical online dating user, Angela has been doing online dating for 4 years. She has had many first dates, only 30% of which turned into a second date. None of those first dates turned into a relationship. After 4 years of online dating, Angela has been extremely frustrated, and found herself unable to connect with the other person in a meaningful way on a first date.
“I don’t know how to emotionally connect with guys on dates…I don’t even know what questions to ask anymore” —Angela
Conversations with Angela
The discovery
The problem with the current online dating system is that the so-called online “dating” platforms do not facilitate actual dates.
Online dating users have trouble establishing rapport with the other person on first dates.
Online dating platforms, such as OkCupid, rely on match questions as an abstract trust proxy to help establish perceived affinity between two online daters. Users spend time and effort inputting their user data into the system by religiously filling out those match questions. However, because their user data remain online as an abstract trust proxy, users do not gain enough value in return.
INITIAL INSIGHTS
A thought came up during the discovery phase: can we bring the match data—i.e. the match questions and users’ answers to the match questions offline, to help online daters establish real connections on a first date?
For research purpose, I decided to focus on OkCupid, a widely popular dating platform that relies heavily on match questions as the matching mechanism. An example of OkCupid match questions:
OkCupid match questions are either multiple-choice or binary questions. Those questions cover a wide range of topics, everything from dating, ethics to sex and politics. Apparently OkCupid takes user’s data and conducts all sorts of research study. According to OkCupid, answers to these three questions predict whether you and the other have long term potential: “Do you like horror movies?” “Have you ever traveled around another country alone?” and “Wouldn't it be fun to chuck it all and go live on a sailboat?” OkCupid analyzed the profiles of nearly 35,000 successful couples who met on OkCupid, and a third of lucky lovers agreed on all three questions.
These are all wonderful hindsights indicating that there could be powerful insights into these match questions and user data. In the current online dating ecosystem, however, these match questions and user data stay online as an abstract trust proxy to help establish perceived affinity between online daters. Users go on dates and forget about these match questions and user data. As a result, the match questions and user data are not utilized to its full potential. I argue that the match questions and user data could be a great conversation material to help users gain deeper understanding of each other and establish better rapport on a first date.
Illustrated by the user journey mapping below, users would fill out match questions during the online matching process and OkCupid would subsequently match two people based on their responses to those match questions. If two people eventually decide to go on a first date, the idea is to reuse those match questions and user data to help daters build real connections offline.
UX Pattern research ON FIRST DATEs
In order to figure out how this idea could help users establish better rapport on a first date, I need to gain a deeper understanding into the UX pattern of a first date.
I conducted a survey on SurveyMonkey.com, and 71 anonymous responses were collected from SurveyMonkey.com’s paid respondent service. Out of 71 respondents, 33 are male and 38 are female.
In the survey, a first date is categorized into five stages: 0% (the moment when two online daters meet in person for the first time), 25% (25% completion into a first date), 50% (50% completion into a first date), 75% (75% completion into a first date), 100% (the moment when two online daters say goodbye at the end of the date). The user experience pattern of a first date is mapped out in terms of the evolvement of emotions, physical contact and conversation topics. The research findings are summarized as below:
User Scenario
The user scenario was created based on extensive research on the user experience pattern of a first date described above. An important premise of this idea is that there is a mutual attraction between two daters. About 75% into a first date when two daters are more comfortable with and have gained some basic knowledge about each other, they are searching for more topics to talk about. OkCupid match questions can help to spice up or hotten up the date and break down physical boundaries at this point — hence the name HotDate.
User Personas
Who could be the potential users of HotDate?
Someone who is shy and has trouble initiating conversations would be an obvious one.
Based on my user research, a lot of online daters mention their first date experiences resemble job interviews. HotDate could create a more playful first date experience for someone looking for a fun first date.
36 Questions to fall in love is a dating conversation starter. HotDate could also target at the same user segment for 36 Questions to fall in love.
Measure of success
Three potential ways of measure of success for HotDate could be:
Fostering emotional closeness
Reducing physical proximity
Creating a playful experience
Competitive analysis
Since the concept of HotDate has been validated, I could move on to develop the actual content strategy.
Before getting into content strategy, I conducted a competitive analysis and found out that 36 Questions to Fall In Love is the only widely-used dating conversation starter. There are currently 16+ live mobile apps using 36 Questions to Fall In Love.
Based on a complete list of 36 questions, the questions are broken down into three different levels and they get increasingly personal as the level progresses. Most of the questions are open ended question addressing second-person pronoun “you,” such as “for what in your life do you feel most grateful?” Open-ended second-pronoun questions could feel a bit interrogating and too serious on a first date.
There is one question, “make three true “we” statements each. For instance, “We are both in this room feeling ... “ addressing first-person pronoun “we” that focuses on commonality, which could be an interesting approach.
On almost all the 36 questions to Fall In Love mobile apps (an example below), all the questions are presented as a list, and users use the arrows to navigate through all the questions. This kind of interaction is quite basic and dull.
Macro interaction design
How can users interact with the match data (the match questions and users’ responses to the match questions) and their devices on a first date?
I explored three possible ways of interactions:
Interaction I: On the first date, dater(s) receives match data in the format of text prompts.
Interaction II: Two daters take turn to send each other a piece of match data with their own device.
Interaction III: Two daters view and discuss their match data on one device.
Three Possible Ways of Interactions Created by Eliana Feng @ Polaris Design.
Rapid Prototyping and user testings
In order to validate the concept of HotDate as well as explore the above three ways of interactions in a cheap and agile way, I needed to figure out how to simulate the above experiences and user test them on 10 real OkCupid first dates.
Interaction I use testings: I collected daters’ match data before they went on dates. Then, I randomly selected a sample of match data and texted them to one of the daters during their first date.
Interaction II user testings: I collected daters’ match data and made them into paper cards, as illustrated below. The paper cards were then divided into two piles, one for each dater. During the first date, each dater would take turns to pick one question from their own pile and discuss it with the other person.
Interaction III user testing: During the first date, one of the two daters logged into their OkCupid profile and review all the match data on one device together.
The discovery
The user testing results show that users normally don’t check their phones during first dates and almost didn't notice the text prompts. Therefore, Interaction I would not work well in real life situation and this experience has to be a human-initiated and mutually consented activity.
Interaction III works better than Interaction II because one device colocates two people, and thus reduces physical proximity.
Proof Of Concept
The design question is: as opposed to an organic flow of conversations, does relying on external conversation materials help enhance the first date experience?
The user feedbacks from the 10 OkCupid dates have validated the concept of HotDate that relying on external conversation starter, or specifically, OkCupid match questions, does help improve first date experience.
The Pivot
As a reminder, OkCupid match questions come with a set of choices, and the original concept is to bring the match questions that users have already answered during the online match process along with their choices to a first date, to help spice up the conversations. However, the technical limitation is that OkCupid does not have public APIs so I am not able to scrape user data from their profiles. As a result, I have pivoted from the original concept, and concluded through user testings that there are still substantial values to the match questions themselves as a conversation catalyzer. In addition, a lot of users did not remember what answers they picked online. Therefore, I have decided to pre-select and curate a set of OkCupid match questions and bring them to first dates where users can pick and discuss their answers on the spot.
Content Strategy
Different from 36 Questions to Fall In Love that are open-ended freeform questions, OkCupid match questions are single-answer questions, asking users to pick just one choice from a predetermined list.
The original match questions, such as "as you get older, what do you look forward to the most" could sound a bit interrogating. Inspired by a widely popular party game Cards Against Humanity, I redesigned the match questions into the Mab Lib format: "As I get older, I look forward to ______ the most," which is more playful as well as encourages self-expression. In addition, I also redesigned some of the questions into guessing questions, for instance, “If you went to my place for the 1st time, and it turned out to be a shithole, I think you would ______.” Through user testings, I found out that guessing encourages eye contacts, and hence could help foster closeness and break down physical boundaries.
Inspired by the theory of Five Levels of Intimacy, the content design follows the principle of “a sustained, gradual, reciprocal and personalistic self-disclosure.” I sorted those redesigned questions into three different levels: "mild," "medium," "hot," encouraging expressions of personal opinions, feelings, and desires, as shown below.
Micro interaction design
Since OkCupid match questions come with a set up choices and users would pick their answers on the spot, what are some of the ways in which users could interact with the choices?
A deeper look into all the match choices, the match questions could be further categorized into: binary, scale and the usual single-choice questions. I created the interaction design for each type of question that could further encourage physical proximity, as shown below. The idea for the interaction designs below is that each dater would control their own avatar (A or B) and move it to the correct position to reveal their desired answer. During the process, their fingers and hands could be touching!
user testing and The discovery
All the above three interaction designs were further user-tested with paper prototypes, and I found out that most users would verbally express their answer before even physically interacting with the choices, ideally shown as above. Therefore, this step of interacting with the match choices seems forced. As a result, I decided to simply let users expressing their answers freely in a verbal way.
Interface Design
When it comes to the actual interface design, I created something simple because the content — the match questions and choices — are supposed to be the hero. The question is presented on the front of a card, and when it is flipped over, a list of choices is on the back of the card for users to freely express their answers. Sequencing the presentation of information, in other words, separating the question and the choices helps bring in an element of surprise and playfulness.
The design of card UI is inspired by Games Against Humanity. Users are already familiar with the card format for games and it is a great way to contain a lot of information, such as presenting the question on the front and the choices on the back. The visual design is inspired by a weather app. The concept for the visuals is to create something playful and bold. In addition, since HotDate is content and text heavy, playing with the typography could make the content itself more engaging.
Live Prototype and user testings
A live prototype was produced with Ionic Framework (Node.JS, Angular JS).
The live prototype was tested with four pairs of couples on their first dates. The first three pairs went on a first date, and the last pair is married. Angela also volunteered in the play-testing. Most responses are positive, further validating the concept of HotDate. Below is Angela’s feedbacks playing HotDate:
“[HotDate] probably saved my date..thank you..[It] was so awkward at [the] beginning” —Angela
Furhter iterations
According to some users:
“[The screen is] also very touch sensitive. We missed a few [questions] because we clicked the wrong way
Therefore, in the latest iteration, a rewind function has been added in case users have accidentally skipped questions.
Users, especially female users mentioned they are uncomfortable with the shaming and hygiene questions, which in the latest iteration have been replaced by questions on moral dilemma and future goals.
However, not every user responded favorably to the idea of Hotdate. Jonathan pointed out that he usually moves fast on a first date and it breaks the flow and sexual tension to play HotDate . His feedbacks helped me discover that HotDate is less suitable for people who tend to move fast on a first date.
“Usually I have a natural vibe going so it would kind of break the flow/sexual tension to play…I also move very fast and usually hook up on the first date” —Jonathan
Therefore, HotDate is designed for users who take it slow and would like to get to know the other person better on a first date.
Market Evaluation
I created a landing page and Google Adwords campaign for a quick market test. The result shows that the clickthrough rate for HotDate’s Google Ad was 3.21%, which indicates a market potential.
deeper Thoughts
Love has always been considered as a complex subject that takes months or even years to happen. The lab experiment of 36 Questions To Fall In Love suggests that intimacy, or even love can be approached in a scientific way. As writer Len Catron who has experimented 36 Questions To Fall In Love on herself commented, “love didn't happen to us. We’re in love because we each made the choice to be” (Catron 2015). Inspired by 36 Questions To Fall In Love, HotDate is intended to continue the discussion about the reality of love. Following a series of scientific approaches, intimacy, or even love could be expedited.
User feedbacks and responses show that people aren’t as receptive to the use of technology on a date as we might expect. A large portion of the people who have been approached about the concept of HotDate believe that love should naturally happen, and the use of technology on a first date could potentially ruin the date. Some of them suggested that HotDate should be made into card game, which in their opinion, is more “natural”. Ironically, all of the people who have been approached are active users of online dating platforms. In some way, they have accepted the idea that love does not need to happen organically — i.e.running into the one in real life, and that technology could be used to find love. Therefore, why can’t technology be used to discover love on a first date? How the relationship between technology and humanity would further evolve in the future would be something interesting to expect.