Statistics from Altmetric.com
Courtship, a key relationship for most adults, has a known association with health.1 Online dating sits alongside traditional relationship brokerage through the family, the church and the workplace.2 Daters wink, right-swipe (on mobile dating apps with a geospatial location like Tinder), email, text and chat online before face-to-face encounters.3 It may take months of browsing and hundreds of invitations before a single response materialises. There are differences in how men and women use the online medium,4 but all users make decisions with limited information. Is there an optimal, evidence-based approach to online dating?
There is much to be learnt from attraction and persuasion research for improving effectiveness of online dating. This literature is scattered across psychology and sociology, as well as computer, behavioural and neurocognitive sciences. We synthesised this evidence to learn how online dating could be improved, maximising the chances of converting electronic communication into a face-to-face meeting.
We performed an extensive search in English language, using Psychinfo, PubMed and Embase in November 2013. Construction of search term combinations was not straightforward as there are no specific indexing terms to target online dating. The search terms we used captured the concept ‘internet OR net OR online’ combined using AND with ‘dating OR love OR courtship’ employing a combination of indexing terms, text words and word variants appropriate to the terminology and structure of each database. We identified further studies by examining the reference lists of the relevant articles selected from electronic databases. We augmented the search further by looking through lists of references generated using the ‘cited by’ function in Google Scholar for each of the relevant articles captured in earlier stages of searching. Study selection followed the principle of saturation5 with the search stopping when no additional studies could be found to address the issue at hand.
We carried out a meta-narrative synthesis,6 an approach suitable for topics that have been studied by different groups of researchers, for different reasons and in different ways. Using primary studies and systematic reviews in psychology, sociology and computer, behavioural and neurocognitive sciences, we sought data for synthesis into broad themes or key messages for various aspects of online dating, drawing on scoping review methodology.7
From 3938 citations captured electronically, 54 were included in the systematic review (figure 1). A further 32 were identified from citation searches and reference lists. We wanted to focus on conversion to face-to-face meetings as the outcome measure as moving merely from online registration to computer-mediated communication lacks the experiential richness2 without which there can be no progress in courtship. Studies did not directly address the question of how online dating contributed towards a lasting partnership. Instead, they focused on individual steps, for example, what features in a screen name or photo increase likeability in the pathway leading to a date. Taking the psychological sciences perspective,8 we outlined the pathway in the literature (figure 2) and sought studies to summarise the evidence for effective strategies at each step (table 1). The range of design features in the literature synthesised included randomisation in 28 studies, cohort follow-up in 13 studies, cross-sectional evaluation in 37 studies, qualitative analysis in 5 studies and systematic review in 3 studies (figure 3).
Creating a profile
Screen name: Desirable names are more often associated with attraction than undesirable names.9 Names with negative connotations (eg, Little, Bugg) are often associated with inferiority.10 Playful screen names (eg, Fun2bwith) are universally attractive. Men are more attracted to screen names that indicate physical attractiveness (eg, Blondie, Cutie), whereas women are more attracted to screen names that indicate intelligence (eg, Cultured).11 Apart from the symbolic significance of names, their alphabetical order plays a role too. A variety of measures of success (eg, educational achievements and income)12 ,13 are correlated with names higher up in the alphabet. Search engine listings are also sorted alphabetically: screen names starting with a letter near the top of the alphabet are presented first. Those in the lower quarter of the alphabet will be lost in the bottom of the pile if you start at the top. Perceptions of similarity-to-self and liking developed as a result of name similarity heighten one's attention and make one more willing to respond.14 Screen names are unique and cannot usually be changed once registered. There is an opportunity to exploit the name-similarity effect by browsing extensively before registration, identifying profiled names of people who you find attractive and then choosing a similar screen name.
Primary photo: In the absence of prior acquaintance, attire and physical appearance in still photos have a powerful influence on likeability.15–17 A genuine smile, one that crinkles up your eyes, will make a good first impression.18 ,19 A slight tilt of the head can also enhance attractiveness.20 Choice of red in a woman21 enhances men’s attraction leading to significantly more contacts. Photo-similarity effect14 may be used in the same way as that described for screen name.
Headline message: Simple language, not over-complicated wording, is likely to result in significantly higher ratings of intelligence because people are naturally drawn to words that are easy to remember and pronounce.22 ,23 It makes information-processing easier, which also increases likeability. Overall attractiveness of the text is positively correlated with photo attractiveness.24 If you can get the potential date to stop and think about your headline message, increasing the exposure time to the primary photo, this will increase their liking.25
Photo gallery: Group photos showing other people having a good time in your company are desirable.26 Women find a man more attractive when they see other women smiling at him.26 Capitalising on the centre-stage effect by selecting photos where you are in the middle creates a sense of importance.27 This can be further enhanced in group photos where you are shown touching another person (confining this to the upper arm to be socially acceptable).28 ,29 This is because a toucher is perceived to be of higher status than the one touched.29 Dynamic video clips can be more realistic than still photos and may promote familiarity at the first face-to-face encounter.30
Description: Individuals are pursued or ignored based on a quick perusal of the profile.31 The prospect of ending with a face-to-face meeting is best met through a profile closer to reality.32 However, it cannot be all about you (bright, fun, non-smoker, into detective novels and long walks on sunny beaches). This will attract far fewer responses than a combination of who you are and what you are looking for. The combination in a 70:30 ratio (genuine, attractive, outgoing, professional female, good sense of humour, into keeping fit, socialising, music and travel, seeks like-minded, good-natured guy to share quality times) achieves the best results.33 Be sure to present character traits but remember that likeability is more important than academic achievement. What characteristics and traits should be revealed? In the absence of familiarity, men prefer physical fitness in women gained via yoga, aerobics and gym, not via rugby and bodybuilding, while women prefer bravery, courage and a willingness to take risks rather than kindness and altruism in their partners.34
How do people make choices in side-by-side comparisons when browsing? They subconsciously check out what is different between those competing for their attention. For example, they may begin to look at who out of those in front of them is most similar to themselves.14 In this situation, people tend not to take into account attributes they would really like to see in their partner.35 A competitive profile would seek to highlight features that distinguish them from other online daters. Adding humour can give the edge (vintage, educated, looking for a woman of ex-film star quality).36 ,37 People assess the cues unintentionally given off (spelling mistakes depict the light on educational attainment) in addition to those cues that are purposefully given. Unintended cues are a powerful way to show without telling,38 for example, a profile written in a humorous, clever manner will be seen as more credible evidence of a sense of humour than the use of the words ‘I am hilarious’. It is also a means of boosting self-esteem.39 ,40 Dishonesty in profiling is damaging to the online dating goal. The profile should have a balance between comprehensive honesty and positive self-presentation because its validity will be put to test in future face-to-face interactions.31
Browsing, invitation and communication
Browsing and selection: Although browsing profiles can itself be a rewarding exercise,41 too much choice can be paralysing.42 One can get into a counterproductive assessment mindset.43 People may browse using essential and desirable traits as search terms. However, side-by-side profile comparisons tend to deemphasise matching against preset selection criteria for the attributes sought in a partner.35 ,44 When the profiles themselves appear genuine and there is a sense of shared identity, a positive feeling46 will lead to a desire to extend an invitation.
Invitation: People almost always see themselves as unique.47–49 Once a group of potential dates has been assembled, a personalised plea is required.50 An individually tailored communication that transforms a cyber-dating digital face into a fully functioning human being is desirable.51 One short, positive remark, directly addressing the person’s character or photo, will do. We routinely reject unrealistically positive views of ourselves,52 because this raises suspicion about the motives of the complimenter. An invitation rhyming with their name or headline message will go a long way, as rhyming poetry has an instinctive appeal.53
Initial electronic exchange: In response to an invitation, a range of different outcomes are possible depending on the quality of communication.54 Compared to a face-to-face interaction, disclosure is far more likely electronically.55 To get started, an open-ended question such as ‘What did you like in my profile?’ is brilliant as it has many possible positive answers.56 Eager responses are not a turn-off, so do not keep the potential date waiting.57 Spontaneous wit and humour58 ,59 puts people in a good mood, even if momentarily, and the way we presently feel exerts a powerful influence on the way we imagine we will feel in the future.60 Disclosure of personal information to each other will make you feel closer.61 Knowing each other’s trivia is a predictor of a long relationship.62 We like others most when we are uncertain as to whether they like us a little or a lot.63 Birth order is informative about personality as younger children tend to be more open, creative, unconventional and rebellious than the eldest child.64
Detecting deception: Concern about being fooled by false information online complicates the process of forming relationships.38 ,65 Sometimes, people present themselves as they hope to be in the future, not as they are at the time of writing their profile.31 People might lie about important issues, such as qualifications or employment; however, they are reluctant to lie in written communication because their words are recorded and may come back to haunt them.66 ,67 Online daters who save early correspondence can check it against later communication for evidence of deception.68 On a video chat, signs of deceit are not those typically associated with increased anxiety; instead, liars more often look as if they are thinking hard for no good reason and to converse in a strangely impersonal tone. They also show behavioural shifts, such as suddenly becoming static and cutting down on gestures.69 ,70
Video communication: We make inferences about emotions by registering body language, listening to voices and noticing facial expressions.71 Making expressive hand gestures during conversation, nodding your head when the other person is speaking, using emotionally charged words (love, like, fond), varying the pitch of voice,72 sitting upright (vs slouching) and smiling (taking over half a second to spread it over your face)20 convey happiness and imply liking.73–75 A smile cheers you up as well.76 ,77 Subtly mimicking movements demonstrates that the chemistry is right.78 ,79
Introducing humour in the conversation makes people more relaxed and accepting.59 People feel greater intimacy when they agree about dislikes more than likes.80 Gossiping positively about others is good.81 Whatever traits you assign to others are likely to be viewed as part of your own personality. Do not criticise, but do not always agree initially.82 It may be better to disagree first, with a view to agreeing later on. Pretending that you are a scarce resource worth having is not universally attractive.83 Mistakes are far more noticeable to us than to others, so if you make one, continue as if nothing had happened.84 If you fear that you are being seen as too perfect, making a small mistake (eg, spilling a bit of coffee on the table and then wiping it clean with a serviette) may actually go down well.85 If you do have a fault to declare, get it out in the middle of the conversation.86
Plan for a positive termination of every chat. We recall items at the end of an experience far better than those at the beginning or in the middle.87 The brain stores a synopsis in which the final scene plays a crucial role.88 We tend to judge the pleasure of an experience by its ending.89 We treasure memories of unusual moments and closing moments.90–92 Reveal positive things about yourself towards the end of a conversation.93 Do not leave the shift from computer mediated to face-to-face communication too late (3 weeks is better than six). Early switch is associated with better outcomes.94
Research results from a range of disciplines synthesised in this review suggest a number of strategies that can increase or decrease the likelihood of converting an online dating site contact into an in-person meeting. These include a huge variety of effective approaches in posting the right photos; writing and rewriting the personal description; browsing for and sending tailored emails; and openly engaging in electronic communication. These may be time-consuming, but there do not appear to be shortcuts in successfully converting electronic contact with innumerable potential dates into a face-to-face encounter with one.
This review has several limitations. First, the search term combination was difficult to construct owing to the lack of relevant subject headings or keywords. Our search may lack specificity, but the iterative process we employed maximised sensitivity. Second, the outcome measures reported in the literature had limitations. A face-to-face encounter and intermediate steps leading up to it are not indicators of a lasting relationship; nonetheless, the long-term follow-up needed for such studies is impractical in real life. Third, data assessment by only two reviewers might limit the validity of meta-narrative synthesis, though this is superior to assessment by a single reviewer. Fourth, there was a wide range of methodological features employed in the studies reviewed, a heterogeneity that is unavoidable when there are a variety of hypotheses addressed with an assortment of research genres. Fifth, the data collated were not amenable to meta-analysis, but this is not necessary in a meta-synthesis. Finally, the generalisability of findings from the literature covering other courtship contexts to online dating needs consideration, though given the evolutionary context described in the paragraph below provides justification for this extrapolation. Despite these features, ours is a comprehensive, robust assessment, uninfluenced by financial interests of online dating service providers, that merits consideration.
Our findings should be interpreted in the evolutionary context. Research on our ability to identify our own emotions shows that we can feel attracted to someone without knowing exactly why.95–98 When it comes to romantic love, we are not as rational as we might think—our limbic system, a part of the brain that deals with emotions,99 overrides or modifies conscious thinking. Developed through ages of evolution, it acts within milliseconds, determining our behaviour instinctively, and in scientific studies it is shown to do so predictably. It directs reproduction and sexuality in all humans, regardless of culture.100 Romantic desire motivates the search for a range of potential partners and romantic attraction narrows it down to specific partners.101 ,102 Courtship energy can then focus on particular individuals and facilitates partner choice.99 ,102 It is no surprise that this powerful mental function, when negatively affected, for example, in romantic rejection, can lead to clinically deranged behaviour including obsession and depression, as well as homicide and suicide in extreme cases.103 Online dating may exploit new technology, but the pursuit of romantic partnership is ancient.
How should online daters employ our findings in practice? Those embarking on this medium can use the evidence-based approach from the outset. Those already registered should update the profile.104 Going public with the plan, for example, telling friends, can provide the encouragement needed to identify the additional time required for this.105 Procrastination can be a problem, and evidence suggests that working on subgoals, even for just a few minutes, can generate the urge to see the task through to completion.106 When bringing things up to date, daters should be mindful that they are dealing with subconscious phenomena,107 and making a good first impression is critical. It may be possible to change the screen name to one that begins with a letter towards the beginning of the alphabet,108 though other things including photos can be updated with features that enhance likeability.109 Remaining close to reality in the update is important because identification of major discrepancies resulting from a conscious desire to disguise the truth will inevitably become grounds for terminating a budding relationship in the not too distant future.110
When ready to launch ahead with invitations, beware that winks (a single click tool to attract the attention of another online member) and generic messages (eg, I like your profile. Would you like to chat?) are impersonal. Potential dates want to feel special. Try poetry (using a rhyming dictionary on the Internet if necessary): For example, if your potential date's screen name is ‘fit n fun’, send a message inviting them to ‘go for a run’; if they are called ‘Iamsweet’, say ‘what a treat’; if they are ‘fitandattractive’, you can be ‘very adaptive’. People positively embrace only those compliments that appear credible, so avoid overt flattery. There are exceptions to these rules in making first contact: where established social norms discourage people from making the first move, winking (a weak signal otherwise) can help achieve matching outcomes.111
In the romantic context, where decision-making is often driven by side-by-side profile comparisons,112 the evidence-based approach will help give an advantage over the competition. Men may have a tendency to try to show themselves in as perfect a light as possible. This can be a mistake, as women may well avoid them, fearing that they are likely to prove attractive to too many others, thereby increasing the competition. Daters should avoid attempts to impress (I have a PhD) or straight appeals for sex. They should use computer-mediated communication as an opportunity to get to know each other before moving to first face-to-face meeting.
Opening an online conversation can be difficult. Closed questions (that have yes or no answers) are bad, as they do not move the conversation on. Open questions (like ‘How are you today?’) that need a full answer are better. With several positive open exchanges, the date is more likely to feel encouraged to move forward.113 Chat about topics that you both find interesting—prepare for these beforehand using the knowledge gleaned from the profiles. Investigate what your potential date does not like. If she does not like beards and you have grown one since uploading your mug shot, shave if off before your first appearance on a webcam. Avoid criticism as negative comments are much more memorable and have a greater impact than positive ones.82 Undoing the damage caused by a negative remark may take more praise than you might think to balance things out. Say positive and pleasant things about your friends and colleagues and you will be perceived as a nice person. Online dating can be used for exploitation.65 As the electronic relationship progresses, reducing uncertainty over any concerns is an important area of computer mediated communication before the first date. Put the specific queries in writing. Be on guard for a sudden reduction in detail in the response and the avoidance of the words ‘me’, ‘mine’ and ‘I’. If someone becomes evasive, press for a straight answer.114
The implications of this review are many. Online daters focus too much on details without realising that likeability springs from subconscious initial impressions. A suitable screen name associated with an attractive still picture and a simple, fluent headline message will immediately and spontaneously generate interest. Choose a screen name without negative connotation that reflects how you would like to be seen by others and that starts with a letter in the top half of the alphabet. Generate a perception of similarity with as many elements of the profiles of people you find attractive as possible. Show, do not tell is the rule of thumb for drafting a profile. The profile should have a combination of who you are and what you are looking for in a 70:30 ratio, staying close to reality and using simple language with humour added. If first impressions are positive, a potential date will move to quickly browse the profile (and then, if sufficient interest is generated, on to a date). If not, they will move on to someone else's profile.
Online daters should browse before registration as nobody picks friends and lovers through random sampling. For the first communication, they should not wink; instead, they should send a short personalised message addressing a trait in the profile of the potential date (try rhyming with their screen name or headline message). They should give genuine compliments, not flatter. When communication gets going and there is a genuine interest, they should start responding soon, keeping individual communication length moderate (to demonstrate generosity with time, but avoid long scholarly theses), saying things with humour (to put the recipient in a good mood), explore common dislikes, and keep the potential date guessing about whether they like them (without playing hard to get). On the webcam, they should mimic body language within reason, gossip positively about others, and end the chat positively. They should get to know the potential date and be prepared to reveal themselves as knowing each other’s trivia is a predictor of a successful outcome. They should move from computer-mediated communication to a date early rather than late.
In conclusion, attraction and persuasion research has identified common pitfalls and effective techniques that can be applied to optimise screen names, headline messages, descriptions of personal traits and electronic communications used in online dating. An evidence-based approach to online dating outlined in this article may provide the key to understanding how to coax the best out of this dating medium.
The authors would like to thank the potential dates who turned down one of us repeatedly, encouraging us to think about the effectiveness of online dating.
Contributors KSK and SC conceived the idea for this paper, searched the literature, appraised the findings, and drafted the manuscript. KSK is the guarantor of the manuscript.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.