Most teams assume that when something goes wrong after a meeting, the issue is alignment.
Someone misunderstood. Someone didn’t listen. Someone didn’t follow through.
But in many cases, nothing “went wrong” in the meeting itself. People were present, engaged and reasonably clear about what was discussed. The problem starts later, when work begins to move forward.
What actually breaks is not alignment. It’s memory.
The same conversation doesn’t stay the same
Ask people to recall what was decided in a meeting a few days later, and you will often hear slightly different versions of the same outcome.
Not dramatically different. Just enough to create friction.
One person is confident a decision was final. Another believes it was still open. A third remembers a condition that others don’t recall at all.
This is not carelessness. It’s how people process information.
Conversations are not stored as exact records. They are filtered through context, assumptions and priorities. Each person walks away with a version that feels correct to them.
Why documentation doesn’t fully solve it
Most teams rely on notes, recordings or AI summaries to prevent this problem.
That helps, but it doesn’t eliminate it.
Notes depend on what someone chooses to write down. Summaries depend on what is considered important. AI-generated outputs simplify by design, removing nuance and compressing context.
All of these are useful. None of them are complete.
They don’t fully capture how a discussion unfolded, how opinions shifted, or how a decision actually formed in the room.
So when teams go back to “what was decided”, they are still reconstructing the past, not referencing it directly.
The gap appears after the meeting
During the meeting, everything feels clear. People agree, move on and switch to execution mode.
The problem appears when work starts.
Someone acts on their understanding. Another questions it. A follow-up conversation reframes the original idea. A small adjustment turns into a different direction.
At that point, it becomes difficult to trace what changed.
Not because decisions are chaotic, but because there is no shared, reliable version of how they were made.
Work doesn’t break loudly. It drifts
In most organisations, decisions don’t fail immediately. They evolve.
Each conversation slightly reshapes the previous one. Each interpretation adds a small variation. Over time, the original meaning becomes less precise.
What looked like alignment turns into divergence.
And because the shift is gradual, it often goes unnoticed until it affects outcomes.
This is not a communication problem
It’s tempting to frame this as a communication issue. But teams today communicate more than ever.
The real issue is that communication is not preserved in a way that supports consistent understanding.
There is no clear, shared reference point for:
- what was actually said
- how the discussion evolved
- when the direction changed
Without that, teams rely on memory. And memory is not designed for precision.
A different way to think about meetings
Instead of asking how to run better meetings, it may be more useful to ask how to preserve what actually happened in them.
Not just outcomes, but context.
Not just decisions, but how those decisions emerged.
Because execution depends not only on what was agreed, but on how people understood it.
Where Ulla fits in
Rather than reducing meetings to summaries, Ulla keeps a structured, traceable record of conversations. It captures the full context of conversations - who said what, how discussions unfolded, and what decisions were based on.
That makes it possible to return to the actual context of a conversation, instead of relying on partial notes or individual memory.
And when needed, this can happen within an organisation’s own infrastructure, giving teams control over how that data is stored and processed.
Most teams don’t struggle because they are misaligned.
They struggle because they are working from different memories of the same conversation.
And without a shared version of what actually happened, clarity doesn’t last very long.