Planning a difficult conversation

A difficult conversation on a project, such as a late delivery, a colleague's poor work or a client's request for work outside the contract, can be planned in advance with robust dialogue, a technique for candid, productive conversations that runs through five moves, from framing the problem to committing to the next steps.

Robust dialogue

Robust dialogue is one of the techniques that can be used to plan a difficult conversation in advance. ^{1, 2} It comes from Insight Experience and is a standard tool used by many people. ³ It allows the writer to conduct candid, open, productive conversations: to gather, understand and reorganise the information to produce robust solutions to the problems, to be open and clear with people so that they say what they mean, and to be creative, come up with ideas, innovate and invent solutions. ²

Conversations that call for robust dialogue

Robust dialogue can be used to plan difficult conversations such as these:

• a statistician saying the tables will not arrive on time
• anyone who was meant to supply information coming back to say they now cannot
• telling a client about a problem with delivery timelines, for example for narratives or for any product
• telling a line manager that the quality of a colleague's deliverables is not acceptable
• a client asking for tasks that are not in the contract and so are not effectively being paid for
• a disagreement about how information should be presented, where the writer is clear that the client's proposal will not work and needs to convince them of an alternative
• a client wanting information in a protocol that is not needed

^{4, 5, 3}

Most writers will have had to deal with one or more of these situations. ⁵

The five moves

The dialogue runs as a circle of five moves. ^{3, 6} From all of them the writer should end up with a productive, respectful, results-oriented discussion. ⁶

• Frame the problem. Define the topic and the boundaries of the discussion so that it does not go off at a tangent, and agree what the dialogue is for. ⁶ A narrow frame focuses the conversation on the single problem, and a problem that is really systemic may need a frame broad enough to cover more than the individual problem. ⁶
• Inquire. Ask the other person why the problem has arisen, drawing out their opinions, reasoning and critical information, and encourage them to be honest about the situation, their experience and their ideas, so that both sides hold as much information as possible. ⁷
• Advocate. Work out what to do, either supporting the situation or recommending something else, and express opinions on what can be done with reasons for how it can be done and why it should be done, communicating the facts and sharing experience to reach a solution that both sides understand. ⁸
• Test and summarise. Check that what is heard is what the other person is trying to say: when they say how the problem will be solved, echo back a summary of the solution, so that both sides give the answer in slightly different words and understand each other. ⁸
• Commit. Both sides commit to the solution by identifying the next steps, who will do what and when, so that clear responsibilities and timelines put actions onto what was discussed. ⁹

A worked example: tables that will not arrive on time

A statistician who says the tables will not arrive on time can be taken through the whole circle. ^{9, 10} Framing decides whether the conversation is about meeting the client's requirements for these particular tables or about a broader problem in the statistician delivering tables on time. ¹⁰ Inquiry asks why the tables will not be available and what caused the problem, such as someone on holiday, sickness or other priorities taking the resource, and whether other staff could be made available. ^{10, 11} It also asks the statistician what they think can be done, and the writer may have suggestions once the reason is known, such as raising the priority of their own work or pulling in people from elsewhere. ¹¹ Both sides need to give their suggestions about what can be done. ¹¹

In advocacy the writer explains why the data are needed, shares their experience of this client and any previous problems, and says why the information is critical. ^{11, 12} The writer may then support the statistician's suggested actions, or give the reasons they think the actions will not work, perhaps with a slight adjustment they think stands a better chance of working. ¹² One possible solution is to ask for the tables in batches, the first five on time and the rest three days later once the writer has dealt with the first ones, and another is that only more help and resource will solve the problem, in which case the writer recommends that the statistician go back to their line manager for it. ¹²

Testing repeats back the solution, for example that the demography tables come first, then the efficacy tables, then the safety tables, with additional resource obtained to provide the safety tables, so that both sides test that it is viable. ¹³ Commitment asks for written confirmation of the revised delivery timelines by email by the next day. ¹³ If the confirmation does not arrive by the agreed time, the writer follows up to make sure there is a commitment on the results of the meeting, and that makes the meeting actually achieve something. ^{13, 14}