Information Revelation in Online Dating

Who thinks you are Hot? Does Knowing Matter?
Information Revelation in Online Dating: A Randomized Field Experiment
Ravi Bapna*, Jui Ramaprasad**, Akhmed Umyarov*1
{[email protected], [email protected], [email protected]}
Background and Motivation
According to the United States (US) Census , 46% of the single population in the US uses online
dating to initiate and engage in the process of selecting a partner for reasons ranging from finding
companionship to marrying and conceiving children, and everything in between. Additionally, recent
research has shown that one-third of marriages in the United States begin online (Cacioppo et al.
2013). Finding the optimal dating and ultimately marriage partner is one of the most important socioeconomic decisions made by humans. Yet, such dating markets are fraught with frictions and
inefficiencies, often leading people to rely on choices made through happenstance, an offhand
referral, or perhaps a late night at the office (Paumgarten 2011).
One of the unique characteristics of online dating markets is that they enable a spectrum of novel
signaling mechanisms that are difficult to replicate in the offline world. Bapna et al. (2014) explore
the impact of leaving a weak-signal through viewing another user’s profile non-anonymously. Their
experiment gifted anonymous browsing to a random sample of users in a given geographic area, and
found that lack of ability to leave a weak-signal resulted in fewer matches. In this paper we explore
the causal impact receiving a strong signal by gifting users a feature that gives them the ability to see
who rated them highly on the dimension of attractiveness. This ability is unique to the online
environment as we explain in detail in the following paragraph. We call this strong signaling feature
‘attractiveness-revelation’ the impact of which we examine within the context of a large North
American online dating site, which we call
On this site, as is typical in online dating, users can receive and provide attractiveness ratings on a
scale of 1(least attractive) to 5 (most attractive). The default setting is that these ratings are secret, in
that they are not visible or known to the users in any form. Rather, they serve as key ingredients for
the background recommendation and matching engine of the site. Thus, the default setting resembles
the offline dating market, in that there exists information asymmetry about people’s preferences for
each other. Specifically, even though a user (ratee) may have been rated highly by another user
(rater), she would not be aware of this. This creates information asymmetry in that the rater holds
information that the ratee does not have access to. The attractiveness-revelation feature breaks down
this information asymmetry by informing the ratee of the identity of the rater if the rater has given
her an attractiveness rating of four or five.
Just as the anonymity feature took away the ability to leave a weak signal, the attractivenessrevelation feature gives the ability to receive a strong signal. We examine the impact of
attractiveness-revelation by asking the following research questions:
1. How does attractiveness-revelation impact a focal user’s mate-seeking strategy in terms of
viewing, messaging frequency and selectivity?
2. How does attractiveness-revelation impact the number of matches (we rely on Bapna et al.
2014 to operationalize a match in the context of online dating)?
3. What is the impact of attractiveness-revelation on a user’s time to get a match?
Author names in alphabetical order. All authors contributed equally. *Carlson School of Management, University
of Minnesota; **Desautels Faculty of Management, McGill University
4. Are time-to-match and mate-seeking strategy differentially impacted by gender and rating
received (four versus five)?
Experiment and Data
In order to test the impact of attractiveness-revelation in online dating markets, we are collaborating
with a large online dating website, (name disguised for confidentiality), and are
currently running a large-scale randomized experiment. Our experiment is being conducted on
100,000 random new users of the website from one geographical area over the period of three
months, which we refer to as month 1 (pre-treatment), month 2 (during treatment) and month 3 (posttreatment). We treat 50,000 of the 100,000 randomly chosen users with a gift of one month of
attractiveness-revelation, making this our treatment group. We keep the remaining 50,000 in the
default setting, such that their attractiveness ratings are kept secret. This serves as our control group
of untreated users.
For each of the 100,000 users, we will have time-stamped viewing and messaging data (both sent and
received), and critically for this study, we will have time-stamped attractive ratings, both sent and
received by these users. Recall, that the default setting does not make these ratings visible to the
users. They are used my monCherie internally to optimize their matching algorithm. In addition, we
have data on a set of demographic variables such as gender (gender = 1 for men, 0 otherwise), age,
sexual orientation (straight = 1 for straight users, 0 otherwise), race. We are also able to determine
whether the users are valid (valid = 0 if the user is a spammer or a bot as determined by internal
algorithms at and we know whether users are active or not. A user is defined as
active if s/he has visited at least one profile ten days prior to the manipulation. In this study we limit
our attention only to users who were valid and active prior to our manipulation.
In summary, our field experiment provides the ability to receive a “strong signal” for the treatment
group while keeping it secret for the control group, allowing us to compare the resulting search
intensity, search diversity, messaging behavior, number of matches and time-to-match between the
two groups.
The experiment is currently running on and will be completed shortly. If given a
chance at SCECR 2014, we will be happy to present the analysis of our results from this data.
Bapna, R., Ramaprasad, J., Shmueli, G., and A. Umyarov. One-Way Mirrors and Weak Signaling in
Online Dating: A Randomized Field Experiment. National Bureau of Economic Research Summer
Institute on the Economics of IT and Digitization, 2013, Cambridge, MA.
Cacioppo, J. T., Cacioppo, S., Gonzaga, G.C., Ogburn, E.L., VanderWeele, T.J. Marital Satisfaction and
break-ups differ across on-line and off-line meeting venues. Proceedings of the National Academy of
Sicences. Available at:
Paumgarten, N., (2011), “Looking For Someone: Sex, Love, And Loneliness On The Internet, New
Yorker Magazine, available at