The article is a part of the Workplace Anthropology Series.

Have you ever met someone, replayed your conversation in your head and thought of the perfect response to a question or statement after the fact? Something that would have demonstrated how witty or knowledgeable you are or how much you have in common? You’re not alone. We’ve all had these moments of what-if. But a recent study reports that many of us may be underestimating how well the conversation may have gone overall. Why is this important? It’s good news for job-seekers, who may agonize over their interview after-the-fact—and it’s an insight into a general anxiety over face-to-face conversations we have, particularly given our increasing reliance on digitally-mediated methods for communication.

While it’s possible that habituation can create a sense of knowing—seeing the same people on your walk to the bus stop or riding in the same car on the train—to know someone, you have to speak with them. Conversation is how we learn about our world and each other. And theoretically a major benefit of in-person communication is that we have access to a host of social cues—eye contact, affirmative nodding, frowning—to tell us how the conversation is going. However Erica Boothby and her colleagues show in many cases we’re ignoring this feedback and are apprehensive about how we were perceived in the conversation. This happens largely due to a fear of social rejection. The behaviors we exhibit are designed to protect our sense of self-worth from a failed connection, but they may also be barriers to making that connection happen. For example:

  • In general, we choose our words carefully but what isn’t said can be as important as what is said. As we replay the conversation and retroactively apply meaning to the silences and phrases that stand out in our mind, we create our own perception of how the conversation went.
  • We tend to hold something of ourselves back until we’re completely certain of reciprocal interest. This means that some of those social cues we’re looking for may appear to be absent because we’re withholding them.
  • Finally, we can miss those social cues when they do happen because we’re so focused on planning for what comes next (e.g., How do we respond? What points do we want to address?).

The researchers have named this tendency the “liking gap.” While initial meetings and relationships at the beginning stages are most vulnerable to these biases, long-term acquaintances are no less susceptible. And interaction between work acquaintances or friends who haven’t seen each other over a period of time can be hedged with uncertainties. Long-standing relationships that have unfolded over time are relatively immune but that is because the relationship has reached a decision point about liking—you either do or you don’t and your exchanges factor in the communication.

Utilizing a methodology orientated around a two-person conversation, the researchers conducted several studies that demonstrate the liking gap in different scenarios:

  • In the initial study, researchers recruited same-sex participants for conversation sessions. The were given ice-breakers and allowed to converse for five minutes before being separated and asked to complete a Likert assessment on their chat. The questions they were asked to evaluate their experience included whether they liked the participant and would be interested in getting to know them better, whether they would want to interact again, how they believed the other participant felt. When mapped, the data revealed that majority of the participants liked their conversation partners, but the majority also felt that their partners did not like them in return. 
  • In a follow-up, the researchers looked for the social signals that we rely on to make these kinds of judgments. There were two scenarios they were looking for. The first is a “no-signal” scenario where people do no outwardly exhibit these signals. The second is a “neglected-signal” scenario where there are signals, but conversationists do not notice or respond to those signals. Third-party coders reviewed the conversations from the initial study and assessed whether the participants liked each other. They were able to predict—based on the behaviors of the participants—how much people liked their partners but unable to determine how they perceived their partner’s response to them. This means that people were visibly providing signals that may not have been interpreted or acknowledged. This supports the idea that we are generally focused on ourselves and can miss important behavioral cues.
  • This focus was validated in a subsequent study that asked participants (in a separate group and allowed to have a free conversation without ice breakers) what key factors went into forming their perception of their partner, and what factors in turn would their partner have evaluated in making a decision about them. When participants reflected on their conversations the factors they emphasized that they believed drove their partner’s assessment were generally more negative than the factors that influenced their perception of their partner. They believed that their partners saw more of the uncertainty and awkwardness that they were feeling than was actually on display.
  • To test the liking gap in longer conversations, researchers used a mix of same-sex and mixed-gender two-pair participants who were allowed to chat for up to 45-minutes. Following the conversation, participants completed a Likert scale survey that assessed their feelings about the conversation and how they believed their partners felt. The liking gap persisted despite the length of the conversation.

Additionally, shyness is absolutely a factor in the liking gap. People who are less shy are less inclined to report a liking gap, whereas people who are shy—even moderately so—do tend to assess a liking gap. 

In the workplace, these results have implications for managers of remote teams, hiring managers, and project-based teams who are meeting for the first time. In instances where there is a clear power-differential, as with managers or interviewers, being aware of these biases means that speakers can take steps to verbally reassure the other participant that they are present, paying attention, and interested. Using statements like “That is an good/interesting point.”, “Can we focus on this statement a bit more?”, or “I like the way you explained that.” can help convey approval where non-verbal behavioral cues can be missed. These types of statements can also help slow the interaction and bring the participant back to specific moment so that they are less focused on figuring out the “right” response or where they might need to add clarity. Finally, simply ending an initial conversation with “I really enjoyed speaking with you” can help take the ambiguity out of the exchange. People are bound to replay the conversation regardless, but these kinds of verbal signals can help reduce some of the associated anxieties from initial conversations.

New project teams may benefit from ice-breaker activities that happen outside of the initial project kick-off. The relationships will unfold over time, but managers of these teams should take the time to bring the group together in a neutral setting so that the focus is less on the success of deliverables and more on the people and the personalities behind the skills at the table.

As we increasingly engage with strangers on assorted digital platforms, as well as build relationships online before bringing them into offline spaces, these findings help us understand why it might be difficult to translate online interactions into offline ones. Online we have the benefit of typing a response, reading it, retyping it a few times and then sharing it. Whether our review happens to this degree is debatable given some of the public exchanges that have played out online but the opportunity exists for consideration. It may also matter less whether strangers online view us as likable if the likelihood of an in-real-life meeting is nonexistent—which may embolden some to speak thoughtlessly or more provocatively than they would in-person.

As we build online and offline networks, we naturally want to populate them with like-minded people. It’s helpful to know that our initial anxieties about these forays into relationships are relatively balanced. 

Comments are disabled here. But you are welcome to join the community on Facebook.


Boothby, Erica J., Gus Cooney, Gillian M. Sandstrom, and Margaret S. Clark (2018). "The Liking Gap in Conversations: Do People Like Us More Than We Think?” Psychological Science

You may also like:

Why did we give our data to Facebook in the first place?

Is email one of the last private spaces online?

What does it mean when we need to take a break from Facebook?

Also in this Series:

The Impact of Politics on Workplace Productivity

Resisting the Depersonalization of the Workplace

Why is cooperation so difficult in the workplace?

Is Quitting Contagious?

What do companies mean by culture?

Using Emotional Intelligence to Crack the Job Interview Code

The Rising Trend of Pets at Work

What can lunch tell us about job satisfaction?

Why are we sleeping with our phones?

Why are we signing our emails with 'Thank you’?

Work and Life in a Hyper-Connected World

Why do we need to have so many meetings?