Whether you’re delegating complex tasks or rallying your team around a new initiative, communicating effectively is an essential leadership skill. Clear, intentional and impactful conversations can not only help you secure buy-in from others but also ground professional relationships in high trust and strong rapport.

The internet is littered with advice on how to pull this off – some based on collections of anecdotes, others on experiments, and still others on frameworks that provide a list of instructions for communicating well. Each of these approaches has its merits, as well as its drawbacks. Anecdotes are vivid but prone to selection bias. Experiments provide evidence for causality but are often separated from real-world contexts. And step-by-step frameworks are structured and easy to follow but tend to rely on generalised principles that aren't tailored to or verified by data from actual conversations. 

What’s lacking is a comprehensive dataset of real-world professional conversations that allows us to identify characteristics associated with effective workplace communication. In this article, we draw on data from a professional services firm that specialises in training and matching executive assistants (EAs) with clients to do just that.

Insights from professional conversations

To date, academics and practitioners haven’t had access to an extensive dataset of naturally occurring professional conversations. To our knowledge, the closest thing available is a large dataset of personal conversations, which differ from professional conversations in norms, goals and style. To fill this gap, we present an analysis of over 2,300 recorded conversations containing more than 1.2 million individual statements. The anonymised data comes from exchanges between leaders – from entrepreneurs to high-level executives – and their EAs.

The professional services firm that fulfilled the placement of these EAs tracked signs of trust and rapport between the leaders and their EAs on a weekly basis. Trust, for instance, was measured through objective, escalating markers – starting from the leader providing their EA with calendar access (low marker of trust); to allowing their EA to correspond with others on their behalf; to providing their EA with access to their personal bank accounts (high marker of trust).

Rapport was defined through escalating markers that tracked the level of closeness implied by the routine communication between leaders and EAs. The lowest marker was the presence of email exchanges – a relatively perfunctory way of communicating. From there, rapport markers escalated to personal messaging apps, phone calls, scheduled one-on-one meetings, unscheduled one-on-one meetings and being a part of the leader’s circle (highest marker of rapport). Studying their interactions, we scored trust and rapport based on whether these behaviours took place between leaders and their EAs. 

We then conducted a regression analysis to test whether elements in the conversations predicted trust and rapport. We used AI to parse and transcribe the recorded conversations, as well as classify them based on whether the statement was professional or personal in nature; if it was about the past, present or future; and whether it was a question. We also counted the number of turns (passing the baton from one speaker to the other) in the conversation. 

Building trust and rapport

Through our analysis, we identified specific characteristics that were associated with higher trust and better rapport:

1. Conversations in which leaders and EAs both participated more in the conversation and exhibited greater turn-taking (i.e. participants spoke one at a time in alternating turns) were associated with greater trust. This may be because greater turn-taking signals greater mutual respect, enhances information exchange and reflects better synchrony between individuals.

2. Conversations with a lower proportion of statements about the past, and a higher proportion of statements about the present and future, were associated with leaders having greater trust in their EAs. This could be because future-oriented conversations suggest proactivity – handling upcoming challenges and problems – which can elevate trust. Psychological research also suggests that people who focus on the future are seen as more agentic and forward-looking, and therefore more trustworthy. 

3. Conversations with a greater proportion of professional statements (relative to personal ones) were associated with better rapport between leaders and their EAs. This might be because strictly professional conversations that place a clear focus on business are more aligned with role expectations of EAs and signal competence and professionalism.

It should be noted that regression analysis, although widely used and accepted, is based on correlations. This means that the findings imply that greater turn-taking, future-orientation and professional focus either cause or reflect greater trust and quality of communication, or both. Regardless, our analysis illustrates what effective professional conversations look like and reveals the behaviours associated with greater trust and better rapport. 

Putting this into practice

Just as people who aspire to be wealthy try to mirror characteristics of affluent individuals and those who aim to be successful athletes emulate the training habits of decorated Olympians, leaders can assess their own professional conversations against these metrics.

Although our dataset covered only conversations between leaders and their EAs, we believe the findings are relevant to a wider range of professional settings. Many workplace situations involve interactions with similar status asymmetries, rapid information exchange and relationship-building, and toggle between administrative details and strategic decisions.

Leaders can facilitate more effective professional conversations that promote trust and rapport among co-workers by cultivating organisational norms. Specifically, we recommend the following:

1. Promote greater participation: Fostering a culture of “making sure everyone is heard” can increase participation during conversations, particularly those between individuals at different levels of the corporate hierarchy (e.g. between a junior employee and a director). Simple norms can help, such as ensuring sufficient time for everyone to speak at meetings, encouraging junior employees to share their thoughts, or prioritising smaller meetings or one-on-ones. Shifting virtual meetings to in-person can also help in this regard.

2. Focus on the present and the future: During feedback sessions with team members, a template that is heavily weighted on the present and future (vs. the past) can promote trust. Similarly, meeting agendas can emphasise current goals and upcoming opportunities with the help of forward-looking frameworks, such as scenario-planning, to shift the conversational focus away from the past. Leaders can also use future-oriented framing in their everyday language, such as “let’s build on where we are now” or “what’s our next step?” instead of “we should have done this differently”.

3. Keep the conversation professional: Although people want to connect with their colleagues on a personal level, our findings suggest that “less is more” in this respect. Designating time for casual chats separate from work-related conversations can help achieve the dual goals of professionalism and connection. For example, leaders can dedicate a brief period at the start of meetings to discuss personal matters – thereby allowing employees to connect with each other – while maintaining focus on work-related issues. This can promote positive perceptions of rapport and an appropriate level of professional closeness. Finally, modelling and perhaps even codifying an etiquette that keeps workplace conversations mostly at the professional level could help.

Effective professional conversations are essential for building trust and rapport in any workplace. By sharing the mic, not dwelling on the past and keeping things professional, leaders can foster a more positive and productive working environment.