Why First Contact Is Not a Relationship
A stranger's urgent email should not become your top priority just because it arrived today. A mistake we found while building Team0.
An unknown person emails you with “URGENT” in the subject line. Your assistant puts it at the top of your morning brief because the message is recent and the sender now appears in your contacts.
That happened in Team0. The email was real. The urgency was real to the sender. Neither meant this person knew the owner, or that the owner should drop everything for them.
The first fix got it wrong too
Our first rule seemed sensible: if the owner had never emailed the sender, mark it as first contact. In an internal check, it marked plenty of people the owner already knew. We did not have a complete record of outgoing mail, so “we found no reply” did not mean “these people have never spoken.”
We changed the check. It now looks beyond email for signs of a relationship: a meeting, a calendar connection, another channel, or earlier interactions. It also checks the owner's own addresses. Without that last step, an owner who wasn't listed in their own World Model could look like a stranger to themselves.
First-contact check
Does the owner know this sender?
A new sender writes to the owner
An urgent subject line is the sender's claim.
Have they met or spoken before?
Meeting, another channel or prior interaction
There is evidence
They may already know each other. Don't call this a first introduction.
Nothing found
Check the owner's own addresses before calling this first contact. Don't turn the pitch into a priority.
This check can still be fooled by a second cold email or an invitation the owner never accepted.
The result is not “ignore this person.” It is “this may be a first introduction.” The assistant can still mention a promising opportunity. It should not let the sender choose your priorities by writing “urgent.”
What we checked
In production data, the new check found first-time senders that the email-only rule couldn't distinguish from known contacts. We also tested it against people with meeting and calendar ties, and people with a substantial history of interaction. Those checks were useful, but they don't tell us how often the rule is wrong across all senders.
There was another problem. The email reached the part of Team0 that writes the brief without the relationship information attached. A better relationship check would have been useless if the brief never saw its result. We had to carry that information through to the email view and make the brief use it.
What still bothers us
A stranger can send a second message or an unsolicited calendar invitation. Either might look like a connection to the current rule, even though the owner has never engaged with them. Requiring a reply or an accepted invitation would be safer, but the records we saw were too patchy to make that the only test without mislabeling known people again.
So the first-contact marker is a warning to be careful, not a final judgment about a human relationship. It certainly isn't permission for an assistant to act.
The larger lesson
Saving an email is easy. Deciding what it means for this person is harder. A pitch can be accurately quoted and still have no place at the top of a morning brief. Someone the owner knows well may have no outgoing email in one connected account.
That is why Living Understanding separates what a source says from the relationship it actually supports. It also asks whether that relationship deserves attention today. The main article explains how those pieces fit together.
Continue reading
Read next
World Model in Practice
Which Promises Are Still Open?
An email that says “sent” is not always proof of delivery. Here's how Team0 handles a promise it has captured as open work.
World Model in Practice
Your Agents Should Not Each Build a Second Brain
Correct one assistant, and the others may still repeat the mistake. Why Team0 keeps one changing picture of the work and gives each agent only what it may see.
Living Understanding
Memento's Problem Was Never the Notes
AI agents are like the hero of Memento: they can't form new memories, so they live on notes. Better learning will help. But the notes still have to stay true.