Abstract
The present study explores the relationship initiation process in online dating using participants’ (N = 105) naturally occurring email messages to a prospective romantic partner. Data were collected online at time one, and participants were recontacted at time two (after meeting their partner offline) to assess the likelihood of continued interaction. A qualitative content analysis uncovered 7 broad categories and 18 subcategories of strategies for initiating relationships in online dating. The analysis indicated that participants’ relationships followed a trajectory that often began with the use of pickup lines to initiate contact and culminated with the transition offline. Along the way, they alternated between strategies for attracting and selecting a partner, constructing an authentic self-presentation, creating a shared context for interaction, revealing and seeking information, and adapting to the online dating environment. Additional quantitative analyses revealed differences in strategy use related to gender and the outcome of the first date. That is, men were generally more direct than women (e.g., by sending the first message), and those who discussed their mate preferences tended to report a higher likelihood of a second date than other participants. The results have implications for the hyperpersonal model, as well as for illuminating the evolution of online dating relationships from the first contact with a partner to meeting offline.
Keywords
Scholars have long been interested in interpersonal attraction and relationship initiation, but the area has received renewed attention in recent years due in part to the rise of online dating (Finkel & Eastwick, 2015; Rosenfeld & Thomas, 2012). Online dating has revolutionized the mate selection process and produced some of the most profound and widespread changes to traditional courtship that have been seen in decades. A survey by the Pew Research Center found that online and mobile dating use among American adults increased by nearly half (from 11% to 15%) in just a 2-year span from 2013 to 2015 (Smith, 2016). Currently, 30% of never-before-married and 19% of previously-married singles report having tried online dating (Anderson, 2016). Popularity aside, online dating also merits attention because of its effects on fundamental interpersonal processes (e.g., self-disclosure, Altman & Taylor, 1973; uncertainty reduction, Berger & Calabrese, 1975). In this way, online dating sites provide an important context for understanding how present-day relationships are formed, as well as for testing extant theories and assumptions about relationship initiation in ways that were not possible during the height of early attraction research in the 1960s and 1970s (Finkel & Eastwick, 2015; Sharabi & Caughlin, 2017).
Online dating relationships do not necessarily follow traditional models of relationship development (Merkle & Richardson, 2000). Rather than meeting a physical person, online daters are introduced to a virtual profile, and much of their early communication occurs exclusively through mediated channels that filter out the social cues people typically rely on to form impressions face-to-face (FtF; Finkel, Eastwick, Karney, Reis, & Sprecher, 2012). This can have the benefit of leading to feelings of enhanced intimacy and hyperpersonal interactions in relationships that remain online (Walther, 1996), but it may also cause setbacks if partners meet FtF and find that their positive impressions were merely idealizations (Ramirez & Zhang, 2007). Accordingly, the present study sets out to explore the relationship initiation process in online dating, as well as its consequences for a relationship’s offline development. Using email messages gathered from 105 online daters, we take a qualitative, inductive approach to uncover the strategies that men and women use to initiate relationships online. We also follow up with participants after their first date to assess whether they were successful in establishing a relationship that is likely to persist FtF.
This study seeks to advance knowledge about online dating and relationship initiation in three main ways. First, we report on participants’ naturally occurring romantic relationships; thus, we add ecological validity to a literature often replete with lab studies (e.g., Wotipka & High, 2016), which, while important, may overlook certain nuances in the ways relationships are initiated and developed online. Second, we avoid much of the bias inherent in self-report methods by analyzing participants’ actual online discourse with a prospective dating partner. In doing so, we observe participants’ relationships from the moment they first begin, and continuing until their first FtF date with a partner. And third, we draw on exceedingly personal, private data that are not readily accessible to individuals outside of the interaction. As far as we are aware, this is one of the first studies to substantively analyze the initial interactions between online daters using digital trace data of their communication behaviors through the site.
Relationship initiation in online dating
Over the past few decades, the Internet has been instrumental in producing large-scale shifts in romantic relationship initiation from FtF to computer-mediated communication (CMC; Rosenfeld & Thomas, 2012). Early theorizing suggested that without the conventional features of FtF interaction, individuals would learn to adapt their communication to the online environment; however, their relationships were expected to take longer to develop due to the slower rate of message exchange and limited social information in CMC (Walther, 1992). More recently, the hyperpersonal model (Walther, 1996) was introduced to explain situations where online communication may even lead to impressions and relationships that are “more socially desirable than we tend to experience in parallel FtF interaction” (Walther, 1996, p. 17). Walther’s (1996) model proposes that CMC relationships can become exceedingly intimate due to the ability for message senders to selectively present themselves and for receivers to fill in the blanks in their impressions by overattributing positive qualities to those they interact with online. Overall, this model would suggest that online dating is not only sufficient for relationship development, but that it may confer certain advantages over traditional FtF courtship (Finkel et al., 2012).
We adopt the hyperpersonal model as a conceptual framework to guide our study of the relationship initiation process in online dating. Despite ample theoretical and empirical evidence that people can form meaningful connections online (Baym, 2015), few have inductively explored the interpersonal dynamics underlying these relationships. Consequently, we work from the ground up to understand how relationships unfold in online dating and the strategies that may position partners for offline success.
Relationship initiation strategies
Research has traditionally examined the relationship initiation process using theoretical frameworks that, unsurprisingly, often predate the rise of Internet-assisted romances (e.g., evolutionary theory, social exchange theory). When applied to online dating, findings from this literature are valuable for uncovering myriad factors that may guide partners’ approaches to relationship initiation, including perceptions of similarity (Tidwell, Eastwick, & Finkel, 2013) and disclosure (Sprecher, Treger, Wondra, Hilaire, & Wallpe, 2013). However, they are not poised to address other qualities that make the act of initiating relationships in online dating unique, such as matching algorithms (Tong, Hancock, & Slatcher, 2016), engagement with pre-interaction profiles (LeFebvre, 2017), or norms for self-presentation online (Wotipka & High, 2016). Thus, it is likely that people use a combination of different strategies when interacting in online dating, making these sites important in their own right, as well as for understanding the communication processes surrounding relationship initiation more broadly.
Recently, scholars have begun studying relationship initiation using online records. Schöndienst and Dang-Xuan (2011), for example, analyzed 167,276 online dating messages to learn what motivates people to respond to contact initiation attempts from potential suitors. Their findings indicated that men were more likely than women to initiate contact, and that messages were more likely to receive a response when they came from physically attractive senders or contained specific linguistic cues. Additionally, Scissors (2010) examined the content of 62 people’s emails for the tactics they used to follow-up with matches after a speed-dating event. Her results pointed to three categories of strategies: (a) ingratiation, (b) future communication and plan making, and (c) text characteristics (e.g., emoticons, exclamation points). She also observed differences related to gender; that is, men tended to be more direct in their advances than women, whereas women relied on more subtle cues to convey attraction than men. We build on this work by exploring online daters’ emails for the distinct strategies they use to initiate relationships. This prompts our first research question:
Gender differences in initiation strategies
It is well established that behaviors in dating relationships can be heavily influenced by stereotypical gender norms and scripts. Conventional norms dictate that men embody agentic traits (e.g., assertiveness), whereas women are expected to display communal attributes (e.g., submissiveness) in their interactions with opposite-sex dating partners (Eaton & Rose, 2011). As a result, men and women tend to exhibit behaviors that cohere with social expectations for their prescribed gender roles, and which can contribute to gender differences documented in past research on relationship initiation. For instance, Finkel and Eastwick (2009) found that the social custom of having men physically approach (rather than be approached by) women at speed-dating events increased their confidence, making them appear less selective in their choice of a mate. This adherence to gendered scripts for interaction may provide structure in contexts such as dating where uncertainty is high (Eaton & Rose, 2011), yet failure to conform may also put people at risk of rejection from potential partners (Brody, 1997).
Despite some initial speculation that the lack of social cues in CMC might make interactions more egalitarian relative to FtF contact (Dubrovsky, Kiesler, & Sethna, 1991), research has mostly disconfirmed this view by demonstrating that many of the same gender differences continue to persist online (see Herring & Stoerger, 2014, for a review). In online dating, men and women still differ in their likelihood of initiating contact (Kreager, Cavanagh, Yen, & Yu, 2014) and reasons for using Internet dating (Clemens, Atkin, & Krishnan, 2015). On the basis of these findings, we also consider the possibility that men and women may utilize different initiation tactics in their online dating communication. Thus, our second question asks:
Online dating and the first offline date
Although much of online dating is about finding someone to date in person (Finkel et al., 2012), less is known about how communication on these sites affects a relationship’s potential for offline success. On the one hand, previous studies generally suggest that the period of online communication preceding the first date should provide a strong foundation that will “remain with the couple when they move to RL [real life] relating” (Baker, 2002, p. 374). If this is the case, then the development of intimate relationships in online dating should portend successful first dates in person. On the other, it is a truism that the first date does not always go well, and hyperpersonal CMC in particular has been shown to lead to disappointment if partners fail to meet each other’s sometimes high expectations after meeting FtF (Ramirez & Zhang, 2007). Thus, a pressing question is whether there are specific ways of initiating relationships in online dating that may predispose partners to better or worse outcomes after transitioning offline. In this study, we focus on the first date as a marker of success because of its importance in “screening out” (p. 1719) potential partners and making decisions about continued offline dating (Whitty, 2008). Our final research question, therefore, explores whether certain relationship initiation strategies are associated with an increased likelihood of FtF interaction beyond the first date:
Method
Participants and procedures
Participants responded via email to an advertisement seeking active members of an online dating platform (app or website), age 18 or older, who were interacting with someone they may meet FtF. The advertisement explained that this was a two-part study, and that participants would be compensated with an entry into a US$75 gift card drawing when they completed part one, along with a guaranteed US$10 gift card when they finished part two. Participants were recruited from online discussion boards, newsletters, and social media sites, as well as offline businesses (e.g., coffee shops, libraries), flyers on university campuses, newspaper classifieds, college classes, and a previous project. Snowball sampling was also used to recruit from participants’ networks.
Qualtrics was used to collect data at two different time points, with time one occurring while participants were interacting with a partner online, and time two taking place after meeting their partner for a first date. Participants received up to eight weekly follow-up emails reminding them about the time two survey. The surveys asked individuals for their demographics, the name of the online dating site they were using, and their likelihood of a second date, along with a host of other questions not pertinent to the current study. Participants were also asked to upload the online dating email messages they had sent their partner, removing any identifiable information. This procedure received Institutional Review Board (IRB) approval, and all participants consented to have their emails used in this study. One hundred and five participants uploaded 207 pages (range = 1 to 18 pages, M = 1.97 pages, SD = 2.42 pages) of typed, single-spaced email data from their online dating history to the Qualtrics questionnaire.
Of those who supplied their emails, 72.4% were women and 27.6% were men. The average age was 24.22 (SD = 7.26), but participants ranged from 18 to 63. They identified as primarily White/Caucasian (73.3%) and heterosexual (90.5%), with nine participants (8.6%) identifying as lesbian, gay, bisexual, or transgender (LGBT), and one unreported (1.0%). Participants’ ethnicities and sexual orientations closely reflect national demographics, with 76.6% of the general population estimated as White (U.S. Census Bureau, 2017) and 4.1% identifying as LGBT (Gates, 2017). Many (48.6%) of the participants used Tinder, followed by OkCupid (26.7%), unreported (9.5%), Match (5.7%), PlentyOfFish (3.8%), other (3.8%), and eHarmony (1.9%). More than half (68.6%) provided data about their first date, and those who did rated their likelihood of a second date on a 0–100% scale, where 0% was “completely unlikely” and 100% was “completely likely” (M = 62.04%, SD = 31.34%).
Data analysis
We conducted a qualitative content analysis (Krippendorff, 2004) of the data in two phases. First, the two authors coded the data independently and drawing on grounded theory and the constant comparative method (Corbin & Strauss, 2008; Lindlof & Taylor, 2019), noted central features of the messages. Through memo writing and conferencing, these emergent details were connected to form strategies (Charmaz, 2006). When analyzing rich or complex data, the concepts researchers bring to the analysis function to bracket the theoretical focus. The hyperpersonal model was used as a sensitizing framework to direct and organize our attention (Tracy, 2013), in this case focusing the analysis on affordances and strategies that may influence relationship initiation online. A codebook was established to define and locate these concepts in the data, serving as a dynamic reference as codes were connected and collapsed, or expanded and given dimension (Lindlof & Taylor, 2019). Three meetings between the authors were used to review and narrow the codes into 7 categories and 18 subcategories of strategies.
Second, we coded approximately 25% of the data for the presence or absence of each strategy. NVivo 11 qualitative analysis software was used to organize the data. After achieving acceptable intercoder reliability (Krippendorff’s α = .71 to 1.0), the remainder of the data was split between authors for coding.
Results
Relationship initiation strategies (RQ1)
The first research question identified different strategies used to initiate online dating relationships. Information that could be used to identify participants was omitted from the quotations; ellipses in brackets denote places where content was removed for clarity and flow. The percentage of participants reporting each of the strategies is displayed in parentheses.
Strategies for initiating first contact with a partner
This category involved one primary contact initiation strategy: (a) pickup lines (cute-flippant, direct, and innocuous).
Pickup lines
Participants who initiated contact with their partner utilized conversational bids that varied in length, effort, and personalization. Cute-flippant pickup lines (7.6%) included statements that were excessively flirtatious or clichéd. As such, they were often generic, making them easy to copy and paste to more than one target. For instance, one man on Tinder’s opening read: “When I get naked in the bathroom, the shower usually gets turned on […] Hey so tell me something about yourself. Favorite food? Color? Your phone number?” There was also an element of humor in this approach, as evidenced by a woman on Tinder who thanked her partner for the “cheesy pick-up line ;).” Past research has shown that cute-flippant openings are not a particularly effective or well-received strategy (Kleinke, Meeker, & Staneski, 1986), which may explain why they were not frequent in our sample. Direct pickup lines (2.9%) were statements making the participant’s interest in his or her partner explicitly known. For example, a man on OkCupid’s opening referenced his attraction to “sarcastic smart asses.” Innocuous pickup lines (23.8%) were differentiated by plausible deniability; that is, where romantic interest was not immediately evident. This was usually some variation of “Oh hey :P How’s it going?” Innocuous openings were by far the most frequently observed type of pickup line, perhaps because they take little effort to compose. Online dating sites can have low response rates, which may incentivize spending less time on each opening gambit. This is demonstrated by a man on OkCupid who wrote: “You’ve already made my 30% reply rate worth it :).” Another possibility is that because relationships can develop slowly online (Walther, 1992), people may be more comfortable using platonic greetings until they have had sufficient time to gauge their own and their partner’s romantic interest.
Strategies for attracting and selecting a mate
This category centered on five strategies for attracting and identifying a partner: (a) displaying resources, (b) relationship alternatives, (c) partner preferences, (d) online dating goals, and (e) matching algorithms.
Displaying resources
The strategy of displaying resources (22.9%) involved participants’ statements about wealth or status, including their occupations, incomes, social connections, or material possessions. For instance, a man on eHarmony referenced his reputation at his former university by noting: “I was apparently one of the most prominent alumni.” Similarly, a woman on Tinder described her career ambitions: “My day job is PR for the hospital, but I have a business that works with personal brands in the sports industry […] I’m in the process of registering as an agent.” Still others drew on their social capital, including one woman on OkCupid who revealed: “I’m out now, picking up VIP passes for the party [NBA athlete] is hosting here :).” Research from an evolutionary perspective has found that resources are one quality women in particular find attractive in men because it signals their ability to provide (Buss, 1988), and the fact that we observed this strategy among both genders may reflect larger societal changes in what men and women value in a mate. That is, as women’s access to status and income has increased, so too has the importance men place on a partner’s financial prospects (Buss, Shackelford, Kirkpatrick, & Larsen, 2001). As a result, women may also display resources as an indication of their ability to contribute equally to the relationship.
Relationship alternatives
The relationship alternatives (12.4%) strategy referred to participants’ dating relationships with others on the site. Some participants noted their positive options outside the current relationship, perhaps to demonstrate their desirability. For instance, one woman on OkCupid said: “I went on a date with a multi-millionaire and some guy who won an Oscar, so I guess it was interesting. Oh, and an FBI agent! Or so they say anyway.” In contrast, several women referenced less suitable alternatives by describing unwanted or inappropriate behaviors they experienced online. One woman on PlentyOfFish wrote: Went out with one guy. It actually went great and we hung out for a week – then he showed up at my house drunk and told me a bunch of crazy/scary stories about his past and I asked him to leave. Ended up blocking him on my phone because he wouldn’t leave me alone. I hid my profile for a week but my friends have met people on here and suggested I give it another go. So here I am.
Partner preferences
The partner preferences (12.4%) strategy included descriptions of the characteristics participants desired in a partner. For instance, a woman on eHarmony said: “I thoroughly enjoy exercising and living an active/healthy lifestyle. I would hope that my partner would want to be active with me :).” Online dating profiles are sometimes written strategically with an ideal mate in mind (Toma & D’Angelo, 2017; Whitty, 2008), and these findings demonstrate how online daters may assess mate value even after identifying a match. Meanwhile, others wanted to know what their partner was looking for; as one woman on OkCupid commented: “I am more confused now than ever trying to figure out what men want. Especially online because there is that disconnect until you actually meet.” Several others referenced qualities they definitively did not want in a partner. For instance, a woman on OkCupid asked several times if her partner had any tattoos, before admitting: “I hate meaningless tattoos :/.” These references to deal breakers show the screening process that some participants followed as they continued getting to know their partner online.
Online dating goals
References to participants’ motivations for dating online were known as online dating goals (10.5%). Some participants talked about wanting to find love, including a woman on OkCupid who said: “I know a bunch of people who have found their husbands/wives online, so I am hoping that it works for me too. Never know…Right?” Yet not everyone was looking for a serious relationship; rather, participants had varied reasons for dating online that did not always involve long-term motivations. For instance, others described using online dating to explore their options, including one man on Tinder who joined “just to see how many matches I could get.” This casual use of online dating was echoed by a woman on OkCupid who described it as her “equivalent of Facebook in that I just look around and waste time on it.” These comments highlight participants’ unique and sometimes unexpected motivations for initiating relationships in online dating.
Matching algorithms
Statements illustrating participants’ attention to the matching process were labeled as matching algorithms (4.8%). Some participants demonstrated a keen awareness of the algorithms used to match them with partners on the site. As a woman on OkCupid wrote: “I like the idea of there being this magic formula of sorts that helps you sort out everyone within a 10-mile radius or so.” Participants also seemed to have a reasonable amount of confidence in these algorithms. One man on OkCupid commented: “At 0% enemy percentage (if we can trust the numbers), we may not like each other, but we’re unlikely to hate each other.” These comments reveal the extent to which the site’s recommendation system may have factored into participants’ judgments of partners.
Strategies for constructing authenticity
This category was defined by the use of two strategies for establishing credibility in online self-presentation: (a) warranting and (b) self-deprecation.
Warranting
The strategy of warranting (11.4%) demonstrated the need to confirm an online identity using outside information (Walther & Parks, 2002), and it involved references to shared networks or requests to view and/or connect with a partner on social media. Although this strategy may be used to escalate a relationship, it can also help accomplish the simultaneous objective of verifying that someone is who they claim to be. Participants routinely asked to add their partner on Facebook or provided their own social media handles. In such cases, this allowed participants to “show” credibility in addition to “telling” their partner who they were through their descriptions and disclosures on the site (Ellison, Heino, & Gibbs, 2006, p. 429). The potential to deceive also makes credibility building, and warranting in particular, a uniquely performed strategy in online dating contexts. After reviewing her partner’s Instagram, one woman on Tinder commented: “You passed. I’m convinced you’re a real person and not a 60-year-old man who wants to keep me in his basement.” Walther and Parks’ (2002) warranting theory proposes that people may search for content that is high in warranting value to connect those they meet online with their offline identities. Consistent with this theory, our participants’ accounts suggest that online daters determine if a partner is trustworthy by seeing if he or she presents a similar image across multiple platforms.
Self-deprecation
Although CMC theories predominately assert that positive self-presentations would be the norm online, participants engaged in self-deprecation (29.5%) frequently. Self-deprecation was used to express a negative view of the self through participant-directed complaints and criticisms. For instance, a man on an unknown dating site wrote: “Dang gurl [sic], you busy. Though, to be fair, it’s common for people to be really busy in the summer. I guess that makes me the boring outlier.” This strategy was also used to offset a positive self-image with negative remarks; for instance, several participants seemed to self-deprecate so as not to appear arrogant (i.e., to “humble brag”). For example, a man on OkCupid said: “Your life is definitely more exciting than mine at the moment! I did manage to land a sweet new job this week so that’s cool.” Such comments illustrate how self-deprecating may help online daters balance the tension between being honest and making a good first impression (Ellison et al., 2006). In this way, using these remarks strategically may develop credibility by creating a sense of authenticity around one’s self-presentation.
Strategies for creating a shared context for interaction
This category included three strategies for creating a “shared social context” (Ellison et al., 2006, p. 430) around participants’ personal histories and online dating experiences: (a) online dating attitudes, (b) perceived homophily, and (c) registering dissimilarities.
Online dating attitudes
A number of participants expressed their online dating attitudes (12.4%) through statements that criticized or complimented the site’s affordances, features, or usability. Some critiques were site specific, while others were about online dating generally. Negative attitudes toward online dating included comments from participants who said they found it harder to write than talk in real time, they thought the process required too much time, they felt restricted by the format, and they questioned the authenticity of the site’s users. For example, a woman on OkCupid wrote: “Confession: I actually hate conversing through OkC [OkCupid]. The profile part? I enjoy writing and introspection. The questions? They’re amusing. But as an email/chat client, eh.” Positive attitudes included comments from participants who said they found it made introductions easier, they enjoyed answering questions about themselves, and they thought the format softened rejection. One man on eHarmony explained: “While eHarmony doesn’t feel like a natural way of meeting people, it definitely helps make introductions to people who have similar interests.” By expressing what they liked and disliked about the online dating process, participants were able to find common ground through shared experiences on the site.
Perceived homophily
As a strategy, perceived homophily (48.6%) referenced participants’ similarities to their partner, including statements that showed agreement or highlighted shared interests, attitudes, or values. Some participants discussed their ideal selves as a strategy for relating to their partner. In this way, partners did not necessarily need to be similar; it was enough for participants to present as if they were using an idealized image. This extends Ellison et al.’s (2006) research by showing how the presentation of an ideal self in online dating may also be used to create shared similarity. For instance, in a conversation about her partner’s child, one woman on OkCupid wrote: “No kids, but have always wanted them. Thought I would have a couple of kids by now but it just hasn’t happened…Life, ya know?” “Bucket lists” are another example of how an ideal self can be expressed online. For example, a woman on OkCupid discussed the places she wanted to visit: “You’ve done some interesting things yourself, traveled around the world a bit. I’ll do that sometime too, eventually Egypt, Greece, Italy, Jerusalem, and Istanbul. And Japan. And the Philippines. But the first five are the bucket list ones.” In this instance, discussing travel interests and future plans helped to construct similarities between partners, even when they did not share similar experiences.
Registering dissimilarities
The strategy of registering dissimilarities (14.3%) was a way for participants to discuss areas of disagreement and friction with their partner. Interestingly, many participants did not appear to differentiate themselves in order to deescalate the relationship, but rather, to reveal information about their more idiosyncratic qualities. For instance, in a discussion about music, a man on an unknown dating site used this strategy to share his own unique taste: “Daughtry’s not my cup of tea, but I’m glad you enjoyed him! I think my top two [bands] for fun would be ‘The Hold Steady’ and ‘Grace Potter and the Nocturnals’.” Other comments about dissimilarity were teasing and flirtatious. For example, a woman on OkCupid joked: “We probably shouldn’t talk anymore since you are a bears fan…Lol.” In each of these examples, attending to dissimilarities was not a way of eliminating a partner, but an opportunity to explore nuances within broader categories of shared interests.
Strategies for revealing and seeking personal information
Participants used three strategies to reveal and request information: (a) incremental disclosures (peripheral, intermediate, and core), (b) resolving uncertainty, and (c) profile referencing.
Incremental disclosures
Participants’ disclosures were organized following the classification of Altman and Taylor (1973). Nearly all participants engaged in peripheral disclosures (90.5%), which were defined as containing biographical or demographic information. Intermediate disclosures (56.2%) were also common and involved semiprivate information such as attitudes and opinions. By referencing other people in these disclosures, some participants suggested what kind of person they were by signaling who they were not. According to a man on OkCupid: “Despite what the writers at the [local newspaper], the mid-50s alumni from the suburbs, or the nearby hick-town inhabitants may think, this isn’t really a sports school.” In refuting the opinions of others, this participant made it clear that he did not identify with local journalists, suburban alumni, or “hick-town inhabitants.” Thus, this technique of disclosure through disidentification appears to be a unique way of revealing personal information about oneself in an indirect manner. Core disclosures (12.4%), which contained private information, appeared in some conversations but comprised a small portion of the disclosures. For example, one woman on OkCupid shared her history of depression, noting: “I am aware that this is one of the themes you have a difficult time talking about even with someone very close, and even more difficult with a picture from the Internet :).” In sum, our results support the general consensus in the literature that CMC is conducive to disclosure (Tidwell & Walther, 2002), yet they also show that most participants avoided oversharing by keeping their remarks consistent with the developmental stage of the relationship.
Resolving uncertainty
Participants expressed ambiguity in a strategy referred to as resolving uncertainty (20.0%). Uncertainty often emerged in participants’ communication as attempts to anticipate and explain their partner’s behaviors (Berger & Calabrese, 1975). For example, a woman on Match predicted the following in anticipation of the first date: “Looking forward to meeting you in person today, though I do think I might be a little too ‘non-city girl’ for you, but who knows.” Others tested the accuracy of their impressions, such as a man on Tinder who said: “So you’re my age and probably from [local] area. Wonder what else I could guess at before an actual meet and greet.” These excerpts also demonstrate how participants gathered information through direct uncertainty reduction strategies (i.e., questions and disclosures), a finding that supports Tidwell and Walther’s (2002) observation that direct strategies are particularly well-suited for information acquisition online. Thus, although the reduced cues online may exacerbate uncertainty, these interactions allowed participants to ask for clarification and seek information from their partner.
Profile referencing
The strategy of profile referencing (23.8%) enabled participants to draw topics of conversation or ask follow-up questions based on their partner’s profile. For instance, a woman on OkCupid said: “I remembered you saying on your profile you’ve done some teaching and you’re a mean teacher! How did you decide to take that approach?” Another woman on PlentyOfFish wrote: “I read your entire profile and you sound like someone I could be interested in. Do me a favor and check out mine (I know, it’s rather long, but it’s pretty accurate) and see if you might be interested.” In addition to the text-based portion of the profile, several participants also attended to their partner’s picture, which resonates with existing research on the centrality of physical appearances in online dating (Schöndienst & Dang-Xuan, 2011). For example, one man on OkCupid said: “Nothing like a picture of backpacking/camping to pique my interest in getting to know someone – especially when there’s a cute girl involved :).” Taken together, these messages demonstrate that profiles continue to be important even after people have made the decision to initiate contact, both as a catalyst for conversation and as a way to learn more about someone.
Strategies for adapting to the reduced cues environment
This category of strategies illustrated two ways that participants adapted to the reduced cues in online dating: (a) temporal qualities and (b) writing style.
Temporal qualities
The temporal qualities (28.6%) strategy involved references to time in participants’ conversations, including statements about response times, delays in responding, or the length of time spent communicating. Like many technologies, online dating affords nearly perpetual contact with partners. In response, some participants added formal starts and stops to their interactions (e.g., saying goodnight to end a messaging sequence) and used their language to control the pace of the conversation, perhaps to prevent the impression that time delays meant their partner was being “blown off.” In terms of how much time had to elapse to be considered significant, one woman on OkCupid commented: “What delay? Two hours is not a delay. Two days is a delay.” Others worried about the kinds of attributions their partner might make based on their response time. For instance, another woman on OkCupid asked: “Did I scare you away by responding too fast? Sorry, you caught me when I was in a responsive mood. I can wait a while to respond to your next message if you like :p.” These interactions illustrate how the asynchronicity of CMC might mean there are ebbs and flows in relationship development, marked by long breaks and persistent streams of conversation.
Writing style
The writing style (7.6%) strategy involved references to composition and message structure, including statements regarding spelling, grammar, or prose. Several participants complimented their partner’s writing or apologized for spelling and grammatical errors, suggesting that the ability to write well may be a unique element of attraction in relationships absent FtF environmental cues. For example, one woman on OkCupid stated: “I like your writing style a lot. It seems settled, well thought of, a style of writing that I rarely get to read in people my age nowadays.” Without FtF indications, written communication can take on greater importance as people rely on small cues for forming impressions and gauging attraction online (Ellison et al., 2006; Walther, 1992).
Strategies for transitioning a relationship offline
Participants implemented two strategies to ease the transition to an offline relationship: (a) expectations for meeting and (b) date requests (direct, indirect, and group).
Expectations for meeting
The expectations for meeting (15.2%) strategy included references to the first date, such as participants’ attitudes and experiences around meeting FtF. Participants placed considerable significance on the first date; for example, one woman on OkCupid revealed her initial reluctance to commit to an offline interaction: “Meeting in person…I don’t know yet. Can I still be ambiguous about it for a while? :))).” Despite being on an online dating site, it was also clear that some people did not necessarily plan to meet anyone in person. Instead, participants seemed to hold different orientations toward meeting offline, as demonstrated by one woman on OkCupid who said: “I’m not bent on meeting anyone in person, but not opposed either. I suppose my only goal is to talk to people.” In contrast, another woman on OkCupid wrote: “I actually meet people from here. I know some people are weird about that.” Taken together, these comments point to two pivotal moments in online dating, with the first being the decision to initiate or reciprocate contact with a partner, and the second being the choice of whether to escalate the relationship offline.
Date requests
Participants formulated their requests for a first date in a variety of ways. Direct date requests (18.1%) were detailed and to the point, including overt statements inviting someone to partake in a specific activity or meet at a certain time. Although a few participants were ready to meet right away, most waited to get to know their partner better. Thus, along with facilitating introductions, online dating also functioned as a central component in the relationship development process. Indirect date requests (22.9%) were passive or vague, including statements that hinted at wanting to meet without making specific plans; for instance, requests to “bring me next time!” This indirectness may allow people to save face if their partner is not ready to meet while still passively signaling interest and delaying specific plans. In this way, tossing the “ball” in another’s “court” puts the onus on the other person to follow up with the request and may be one reason why individuals sometimes encounter difficulties when escalating their relationships offline. Group date requests (4.8%) were less typical and involved third parties and face-saving strategies inviting the other person to tag along on an activity. As one man on OkCupid wrote: “You’re free to come with us if you like? You could invite some of your friends as well?” Unlike other requests, group dates allowed participants to take the pressure off a one-on-one meeting by including their partner in their usual routines.
Initiation strategies by gender (RQ2)
Chi-square analyses were used to explore whether there were gender differences in participants’ strategy use. Prior to the main analyses, we examined which gender typically initiated first contact. Consistent with past research (Kreager et al., 2014; Schöndienst & Dang-Xuan, 2011), men were more likely than women to send the first message, χ2 (1, n = 105) = 25.85, p < .001 (men: 72.4%; women: 19.7%). From there, we probed the data for gender differences in whether participants displayed each strategy in their emails (0 = no and 1 = yes). Women were more likely than men to evidence registering dissimilarities, χ2 (1, n = 105) = 3.84, p = .05 (men: 3.4%; women: 18.4%), whereas men were more likely than women to reference cute-flippant pickup lines, χ2 (1, n = 105) = 9.72, p = .002 (men: 20.7%; women: 2.6%), innocuous pickup lines, χ2 (1, n = 105) = 9.75, p = .002 (men: 44.8%; women: 15.8%), and direct date requests, χ2 (1, n = 105) = 4.52, p = .03 (men: 31.0%; women: 13.2%).
Initiation strategies by first date outcome (RQ3)
Independent samples t-tests were used to examine whether participants’ strategy use was related to the outcome of the first date. Participants who discussed their partner preferences reported a higher likelihood of a second date (M = 88.00, SD = 16.43) than others (M = 59.88, SD = 31.39), t(63) = 1.97, p = .05, and those who cited their relationship alternatives anticipated a lower likelihood of a second date (M = 38.16, SD = 33.83) than other participants (M = 64.47, SD = 30.34), t(63) = −2.00, p = .04.
Discussion
The present study is among the first attempts in the literature to qualitatively analyze the online dating process using digital traces of relationships as they naturally unfold over time. As such, it adds to an emerging body of research that has used email records to capture and understand the real-world dynamics of interpersonal relationships (Schöndienst & Dang-Xuan, 2011; Scissors, 2010). Results of a qualitative content analysis surfaced 7 inductively derived categories and 18 subcategories of strategies for initiating online dating relationships (RQ1). Furthermore, quantitative analyses indicated that participants’ strategies were related to their gender (RQ2) and their likelihood of a second date after meeting in person (RQ3).
Summary of results
Our results for RQ1 illuminate how a number of fundamental relationship initiation strategies are uniquely manifest in the online dating context. To initiate first contact with a prospective dating partner, participants used conversational bids that cohere with the types of pickup lines (cute-flippant, direct, and innocuous) identified in past research on FtF courtship (Kleinke et al., 1986). Participants also revealed and sought information through strategies that largely mirror traditional models of relationship development, including incremental disclosures (peripheral, intermediate, and core; Altman & Taylor, 1973) and resolving uncertainty (Berger & Calabrese, 1975). Additionally, online dating makes it possible to selectively self-present oneself to potential partners (Walther, 1996), and even though participants used strategies such as displaying resources and relationship alternatives to construct a desirable self-presentation, they were also able to offset an overly positive image and confirm their identity through self-deprecation and warranting (Walther & Parks, 2002). Together, these findings are valuable for uncovering a variety of relationship initiation strategies that resonate with past research, as well as for demonstrating how some of these tactics may be enacted differently in online dating.
The results of RQ1 also spotlight the ways different relationship initiation strategies were adapted in response to features and affordances of the online dating environment. For example, matching algorithms played a significant role in how some participants attracted and selected a partner (Tong et al., 2016), while profile referencing supplied information in relationships even after they had progressed past the initial introductory phase (LeFebvre, 2017). There were also participants who commented on their partner’s writing style and the temporal qualities of the interaction in ways that allowed them to demonstrate interest through the use of specific channel characteristics. This is consistent with research showing that without the nonverbal and social cues that people rely on to form impressions FtF, subtle details such as an individual’s writing and response time can take on greater meaning (Ellison et al., 2006). Overall, these findings provide an inductive view of a constellation of strategies that characterize the experience of initiating a relationship in online dating and point to intriguing opportunities for future research.
Additionally, the results of RQ2 underscore how the relationship initiation strategies identified in this study may be influenced by the gender of the initiator. In general, men were more direct in their approach to relationship initiation than women (Scissors, 2010), as evidenced by the fact that they were more likely to initiate first contact, as well as to reference cute-flippant pickup lines and direct date requests in their online dating messages. Men also used innocuous pickup lines more than women, which, while not necessarily direct, still required them to make the first attempt at opening a conversation. Meanwhile, women were more likely than men to show signs of registering dissimilarities, perhaps because they were approached more often and could therefore be more selective when choosing someone to communicate with on the site. These results support a wealth of research on the different mating strategies that men and women use to initiate relationships (see also Eaton & Rose, 2011). Furthermore, they suggest that despite the potential for CMC to support less gendered relational dynamics, men and women may still deploy slightly different strategies when initiating relationships online (Scissors, 2010). Contrary to past research, however, men and women did not differ in their use of other strategies, such as boasting about their status and resources (Buss, 1988) or referencing their goals for dating online (Clemens et al., 2015). Thus, despite finding evidence of some gender differences, it is also important to note the many ways participants were similar.
Finally, our data for RQ3 reveal how these strategies for initiating relationships may be related to the outcome of the first date. Experimental studies of CMC have shown that the transition offline may pose challenges for relationships (Ramirez & Zhang, 2007), yet it is also a necessary precursor for further development. On average, the participants in our sample who met their partner FtF were largely successful in establishing a relationship they thought would result in a second date. We did, however, still observe that some initiation strategies were associated with better first dates than others. Specifically, participants who reported a stronger possibility of a second date were less likely to use the site to discuss their relationship alternatives, and more likely to talk about their partner preferences. In sum, these data provide additional evidence of the potential for a couple’s online communication to spill over to affect features of their offline relationship (Baker, 2002).
Theoretical implications
This study advances the hyperpersonal model by (a) providing a rich description of relationship initiation and development in the context of online dating and (b) highlighting the intersections between existing relationship initiation strategies and distinct features (i.e., reduced cues, asychronicity) of the online dating environment (Walther, 1996). Online dating relationships do not necessarily follow the same trajectory as traditional courtship; rather than beginning with the first date, development on these sites often culminates with the decision to meet FtF. Indeed, our data suggest that one of the primary purposes of the period of online communication between the first email and the first date is to make a judgment about continued dating. Thus, while a number of studies have been devoted to understanding whether CMC’s hyperpersonal effects can be maintained once partners meet FtF (Ramirez & Zhang, 2007; Sharabi & Caughlin, 2017), an equally important question may be why some partners never meet at all. Additionally, our findings may be useful for extending traditional theories of relationship development (e.g., uncertainty reduction theory, social penetration theory) to CMC by illustrating how individuals enact basic interpersonal processes in their online dating communication. Future online dating research utilizing a relational stage approach may also build on this work by exploring the different phases of development implicit in our results.
Limitations and conclusion
This study had several limitations. First, online dating usage tends to peak in middle age (Rosenfeld & Thomas, 2012), meaning our relatively young sample (M age = 24.22) may not fully capture the experiences of older adults. Second, all participants expected to meet their partner offline and more than half did, which suggests that many of these relationships were already largely successful by online dating standards. Third, because participants self-selected into our study, the content of their emails may have been different from those who declined to participate. Fourth, to obtain IRB approval, we only asked participants to provide their own messages (and not their partner’s). Finally, despite our efforts to recruit a diverse sample, our participants were relatively homogenous in regard to demographics; future efforts should be directed at exploring how different categories of users initiate relationships in online dating. Despite these limitations, this study offers new insight into the process of initiating relationships on online dating sites, and suggests that the outcome of the first date may depend on the communication prior to meeting.
Supplemental material
Supplemental_PDF - From first email to first date: Strategies for initiating relationships in online dating
Supplemental_PDF for From first email to first date: Strategies for initiating relationships in online dating by Liesel L. Sharabi and Tiffany A. Dykstra-DeVette in Journal of Social and Personal Relationships
Footnotes
Authors’ note
An earlier draft of this manuscript was presented at the 2018 Conference of the International Association for Relationship Research, Fort Collins, CO, USA.
Acknowledgment
The authors would like to thank Keith Berman and Alexandra Mages for their assistance with data collection, as well as the editor and three anonymous reviewers for their valuable feedback.
Funding
The author(s) received financial support for the research, authorship, and/or publication of this article: Support for this study was provided by the Ruth Anne Clark Award Fund at the University of Illinois at Urbana-Champaign.
Open research statement
As part of IARR’s encouragement of open research practices, the authors have provided the following information: This research was not pre-registered. The data used in the research are available within the limits of participant confidentiality requirements. The data and materials used in the research are available upon request by emailing
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
