The Meeting That Never Actually Happened: Why Retrospectives Fail to Create Contact

The retro that worked on paper

Two years of retrospectives. Action items written and owners assigned. Nothing ever changes. The format has been switched four times. The problem is not the format.

A delivery team at a fintech. Eighteen months of fortnightly retrospectives, each one facilitated with care. Formats rotate — sailboat, four Ls, start-stop-continue. Sticky notes are genuine. Action items are specific, assigned, and dated. The team participates fully. And sprint after sprint, the same three themes appear — phrased differently, categorised differently, voted on differently — without any of them moving. The coach reviews the session notes from the last six months. The facilitation quality is good. The room was safe. The process worked.

Nothing changed.

The problem was not in the retro format. The retro format was doing exactly what it was designed to do: produce a well-managed process. A well-managed process and a genuine encounter are not the same thing.

"The team needs more safety." "The format needs changing." "People aren't being honest." All of these diagnoses are available. All of them miss the point. You can have full safety, a thoughtful format, and complete honesty — and still produce a process in which nobody is actually affected by what anyone else says. That is the specific failure this article is about.

The retrospective format is a container. What goes in the container is not technique but genuine contact — the experience of two or more people actually meeting each other and being changed by the encounter. You can optimise the container indefinitely without changing what is in it.

What intersubjectivity is, and why it matters for retros

When two people are genuinely talking — not reporting, not exchanging information, but actually meeting — something emerges between them that neither of them brought in. A developer says something about the sprint. The tech lead hears it, and their face shifts slightly. The developer sees the shift, and says something they had not planned to say. Neither person predicted this sequence. Neither controls it. The exchange itself is generating something that was not in the room before they began.

This is not mystical. It is the ordinary experience of conversation when it is genuinely alive rather than managed. Most retrospectives are not alive. They are well-organised. The distinction is consequential: in a well-organised retrospective, nobody is genuinely affected by what anyone else says. In a genuinely alive one, people leave having encountered something they did not expect — and that encounter is what changes them.

This space — co-created, irreducible, shaped by both parties but owned by neither — is what Van Beekum (2012), following Stolorow, calls the intersubjective field: "constituted by the reciprocal interplay between two (or more) subjective worlds." There is also, in Parlet's formulation via Summers and Tudor, "a shared field, a common communicative home, which is mutually constructed." The plain translation: the meeting between people produces something the process cannot produce, because the process is not between people — it is between people and a format.

The diagnostic question for any retrospective: is anyone in this room being genuinely affected by what anyone else is saying? If yes, something is alive. If no — if everyone is managing the process, categorising, voting, noting — nothing is happening. The notes are being taken. The meeting is not.

Most retrospective formats function as contact barriers. The sticky note process, the dot voting, the categorisation exercise — each element reduces the risk of someone being actually moved by what another person says. Which is precisely the risk that makes change available. The format provides safety from genuine encounter. Safety from genuine encounter is safety from change.

Van Beekum (2006), quoting Polster and Polster: "Once an organization is in full contact with internal and external process, there is no need to chase change, because it simply occurs." Change is a byproduct of genuine contact. The retrospective format is designed to produce change directly. This is the design error.

What genuine contact looks like in a retro

Contact has specific signs. So does its absence. The practitioner who can read these signs in real time can choose different moves than the one who is managing the agenda.

When contact is happening

Something is said that has not been said in this room before — not because the format created space for it but because the room itself made it possible. Someone changes their mind during the session — not capitulates to pressure, but genuinely updates their position in response to what they heard. A relationship between two people shifts visibly: they look at each other differently at the end of the session than they did at the start. An unnamed theme surfaces and gets named, and the naming feels like relief rather than diagnosis. The session runs over because people are reluctant to end what is happening.

When contact is not happening

People give well-formed answers that feel rehearsed. Action items appear quickly and feel disconnected from the room's energy — generated by the format rather than by genuine understanding. Nobody changes position during the session. The SM knows, within the first ten minutes, exactly what everyone will say before they say it. The retrospective ends on time, action items distributed, and everyone feels a mild, almost imperceptible relief that it is over.

This is not a story about facilitation skill versus its absence. Highly skilled facilitators can run retrospectives that produce zero contact — because their skill goes into managing the process beautifully, which inadvertently reduces the space for messiness. Messiness is often where contact lives. The facilitator who keeps everything clean and efficient may be, in that same move, keeping everything empty.

Contact cannot be forced. But the conditions that make it available can be created deliberately. None of these practices require a format change. Each is a different quality of attention brought to the ceremony that already exists.

Slowing down before speaking

Contact requires slightly less speed than process. The most immediate intervention available to any facilitator: after someone speaks, wait longer than feels comfortable before the next move. Ten seconds. Fifteen. Watch what happens in the room while the words are still present. Most facilitators fill silence within five seconds. The space between five and fifteen seconds is where contact attempts to emerge. The format stays intact. The pace is the variable.

Prioritising what has landed over what is next

The facilitator's process instinct is to manage forward: note the insight, categorise it, move to the next topic. Contact is created by moving sideways: Before we move on — what was it like to hear that? The content slows. The relationship deepens. This is not a therapy question. It is the difference between the team processing information and the team experiencing something together. The first produces action items. The second produces change.

Process reporting — the SM's own experience as retro content

In Van Beekum's consulting seminars, the practitioner group reported to the client not just what they had learned but how it felt to learn it — the content and the relational experience of encountering the content, offered together. For retrospectives: the SM brings their own experience of the sprint into the room alongside the team's. I kept feeling pulled to follow up on the integration issue all week. I'm not sure whether that was about the issue or about something else. This normalises subjective experience as retro content. It also models the kind of presence that makes contact possible — and the retrospective is asking the team to do exactly this.

Following digression rather than redirecting it

The most important retrospective conversations often begin as digressions. Someone mentions something slightly off-topic. The facilitation instinct is to redirect. The contact instinct is to stop: That felt important. Tell me more. Digressions frequently contain the session's actual subject — the thing the team has not found a way to put on the agenda because the agenda does not have a category for it. The format has categories. The team's experience does not always fit them.

The "what happened between us" questions

The genuinely relational question in a retrospective is not "what happened with the work" but "what happened between us while we were doing the work." Three questions that open this territory:

"When was the moment in this sprint when you most wanted to say something you didn't?"

"When did we work best together — not just efficiently, but actually together?"

"Is there something between us right now, coming into this retro, that needs to be named before we look at the sprint?"

These are not therapy questions. They are relational questions — and the retrospective is, by its own design intent, a relational ceremony. The format does not provide a category for them. That is not an argument against asking them.

The role of silence

Genuine contact often requires a pause long enough that the managed-process instinct becomes uncomfortable. Most facilitators fill silence within five seconds. Letting it run to fifteen or twenty is one of the most contact-generating moves available — because it creates a gap the room itself must fill, and what the room chooses to put in that gap is almost always more interesting than what the facilitation plan would have provided next.

What this changes about retro design

Contact does not require abandoning retrospective structure. It requires designing structure to support encounter rather than substitute for it.

Each layer of process complexity — additional voting rounds, multi-stage categorisation, pre-session async work — increases the process management load on the facilitator and team. Every unit of attention spent managing process is unavailable for genuine encounter. The simplest formats are often the most contact-permitting — not because simplicity is a virtue in itself, but because complexity fills the space where contact could emerge.

Begin the retrospective with a question that asks for the person's state rather than their perspective. How are you arriving today? What's present for you as we start? These are not therapy prompts. They establish that subjective experience is retro-relevant data — which is the precondition for anything intersubjective to happen. A team that has not been invited to be present as people cannot be expected to encounter each other as people.

End not with "what are our action items?" but with "what happened between us today that's worth knowing?" This question treats the retrospective itself as an event — something that occurred between people — rather than as a process that produced outputs. It creates the opportunity for the team to name something about the encounter rather than only about the sprint.

There is one caveat. Teams that have no relational foundation — newly formed, or carrying acute unprocessed conflict — sometimes need format scaffolding to function at all. Contact cannot be forced when there is no basis for it yet. In these contexts, retrospective structure serves its appropriate function: providing enough safety for people to be in the same room productively. The error is applying this scaffolding universally — as if every team at every stage of development needs the same protection from genuine encounter. Most established teams do not. They have been protected from genuine encounter for years, and it shows in their action item completion rates.

Your own presence is the variable

The SM's quality of attention shapes what is possible in a retrospective more than any format does. A genuinely curious facilitator who is fully present — not managing the process from behind the agenda — creates more contact than the best-designed format in the hands of someone who is steering rather than travelling.

The development question for the practitioner is not "which format should I use next?" It is: am I managing this team or meeting it? The distinction is felt more than reasoned. Managing feels like steering — you know where the session is going, you are keeping it on course. Meeting feels like travelling together without knowing exactly where you will arrive. Both involve skill. Only one involves genuine presence.

When the facilitator brings their own experience of the sprint into the room as a participant rather than as a process manager, the team encounters a person, not a role. This is exactly what the retrospective is asking the team to do: show up as people rather than as functional roles reporting on a process. The facilitator who models this makes it available. The facilitator who does not cannot require it.

This series began with a Scrum Master finishing a developer's sentences in a sprint review — not as a choice, but as a pull the system exerted and she followed without noticing. The series ends with the same argument from the other direction: the coach's presence is not neutral. It is either absent behind a format, or genuinely in the room. When it is genuinely in the room, the retrospective becomes something other than a managed process. It becomes the meeting it was always designed to be.

The meeting the series has been describing

Six articles. One argument, approached from six directions: the coach is inside the system, not above it. Their behaviour is data. Their history interacts with the team's. The group's defences move through them — through splitting and through projective identification. Their quality of attention determines what becomes visible. In each case the argument is the same: the coach's own experience — their involuntary enactments, their projective identifications, their pre-loaded diagnoses, their managed-process presence — is not the noise surrounding the coaching work. It is the work.

The retrospective is the ceremony where this argument is most testable. It is the moment designed specifically for genuine contact. If the facilitator is genuinely present, something becomes possible that was not possible before. If the facilitator is managing the process from behind the format, nothing moves — regardless of how skilfully the process is managed, how thoughtfully the format is chosen, or how safe the room feels.

Van Beekum (2012): "The reality created by client and consultant is not about objective truth but about intersubjective relating and understanding." The retrospective that never quite happens is the one where this creating never takes place — where the format runs and the people attend and the action items are generated, but nothing between the people in the room is actually at stake. The retrospective that happens is the one where someone is moved by what someone else says, and that movement becomes the data the team works from.

The format is not the meeting. The people are the meeting. The format is just what you do until the meeting starts.