Safe-to-fail Conversations Using conversation as a probe when the way forward is uncertain

In complex situations, we cannot know in advance what a conversation will reveal or where it will lead. Rather than relying on one large conversation to find the answer, we can hold many small conversations in parallel and learn from what emerges. Safe-to-fail conversations allow us to explore possibilities without requiring predetermined outcomes.

Conversation as a form of intervention

We often think of conversation as something that happens around change. We talk to understand a situation, discuss what should be done, communicate decisions, and reflect on what happened. But conversation can do more than support change.

A conversation changes the pattern of interaction between people. A new question is asked, someone who has remained silent speaks, two people who rarely interact exchange perspectives, an assumption is challenged, or an idea is voiced before it is fully formed. Something has changed, however slightly.

If interaction is the primary unit of change in a complex human system, such moments matter. A conversation can itself be an intervention into the system. When that intervention is deliberately small and exploratory, and the consequences of failure are contained, we can think of it as a safe-to-fail conversation.

Conversation as a probe

A safe-to-fail conversation may simply be intended to discover more. We might bring together people who do not normally talk to one another, ask a question that has not previously been asked, invite a perspective that is usually missing, or explore an uncomfortable issue without requiring anyone to decide immediately what should be done about it.

We cannot know beforehand where such a conversation will lead. That is partly the point. We create the conditions for the conversation and pay attention to what happens.

What surprises us? What do people respond strongly to? What assumptions become visible? Where is there energy or resistance? What new possibilities appear?

The conversation probes the system and gives us information that was not available before.

Many conversations in parallel, not one big conversation

Safe-to-fail probes are not normally carried out one after another until we find one that works. In complex situations, several different probes can be run in parallel so that we can see how the system responds to different interventions. The same principle can apply to conversation.

Instead of bringing everyone together for one large conversation intended to produce the answer, we can initiate many smaller conversations in parallel. Different groups might explore different questions. People from different parts of an organization might look at the same situation from different perspectives. Some conversations might bring together people who normally work closely together, while others connect people who rarely meet.

We are not looking for the conversation that gets the answer right. We are looking for patterns. Some conversations may lead nowhere, some may expose difficulties we had not anticipated, while others may reveal possibilities, connections, or sources of energy worth exploring further.

We can amplify what appears promising, pay attention to what is causing concern, and initiate further conversations based on what we have learned. Rather than placing the burden on one large conversation to determine what should happen, we create many opportunities for the system to reveal itself.

Ideas can also be safe to fail

There is another sense in which a conversation can be safe to fail. Good conversations allow us to say things that are not yet fully formed.

I’m not sure about this, but…
What if we looked at it another way?
There is something here that doesn’t quite make sense to me.
This may be completely wrong, but…

These are small conversational probes. We put an idea into the space between us and see what happens. Someone may build on it or question it. It may lead somewhere unexpected, or it may simply disappear because nobody finds it useful. That is fine.

If every contribution has to be polished, defensible, and correct before it can be spoken, much of the exploratory potential of conversation disappears. A safe-to-fail conversational space allows tentative thinking, disagreement, uncertainty, and unfinished ideas. It gives us permission to explore without requiring every thought to become a proposal or every conversation to produce an outcome.

Safe-to-fail does not mean consequence-free

Not every conversation is safe. Questions can cause harm, power differences matter, and people can be embarrassed, excluded, threatened, or exposed. A conversation that feels exploratory to one person may feel dangerous to another.

Safe-to-fail therefore does not mean that failure has no consequences. It means that the consequences of failure are contained and are not catastrophic. Before initiating a conversational probe, we should consider what might happen if it goes badly, who might bear the consequences, and whether those consequences are acceptable.

A tentative discussion among trusted colleagues is very different from publicly challenging someone’s competence or reputation. The possible consequences are different, and so is the degree to which the conversation can reasonably be described as safe to fail. The same principle applies to conversations as to any other safe-to-fail probe: failure should allow us to learn without creating unacceptable harm.

Conversation and the adjacent possible

Safe-to-fail conversations also connect with the idea of the adjacent possible. We rarely see all the possibilities available to us from the outset. A conversation may reveal something that nobody had previously considered, and that idea may lead to another conversation, which opens another possibility.

We discover the path by moving along it. A conversation does not have to tell us where we will eventually end up. It only needs to help reveal what might be possible next.

Direction without a predetermined outcome

This also connects with Vector Theory of Change. In a complex situation, we can have a sense of direction without specifying a fixed destination. Conversations can work in the same way.

A conversation can have purpose and direction without requiring a predetermined outcome. We may know broadly what we are trying to move towards, while remaining open to what emerges along the way. This is explored further in The Vector of Great Conversations.

Safe-to-fail conversations and Conversational Leadership

Safe-to-fail conversations are one way of putting Conversational Leadership into practice. Instead of assuming that we already understand the situation and need only persuade others to act, we approach conversation with curiosity.

We ask rather than tell. We explore rather than prescribe. We listen for what is unexpected rather than only for confirmation of what we already believe. We allow ideas to develop through interaction rather than insisting that they arrive fully formed, and we recognise that no single conversation needs to carry the burden of producing the answer.

In complex situations, change emerges through many interactions over time. Each conversation can make a small difference to the pattern of relationships, understanding, and possibilities within the system. Some conversations will lead nowhere, some will surprise us, and some may change the direction in which we are traveling.

What matters is that we pay attention to what emerges and allow what we learn to influence what we do next.

We cannot talk our way to the perfect plan when the situation is complex and uncertain. But we can create many small conversations in parallel, learn from what emerges, and use those insights to decide what to try next. Each conversation can open possibilities that were not visible before.