Relationship strength, explained
How K3X decides who on your team knows a contact best, which signals count, which ones we ignore, and why the score is a word rather than a number.

K3X Team
Share

Before you email a prospect, the most useful thing to know is whether someone on your team already knows them. A warm introduction from a colleague who has met the buyer three times beats a cold email from you every time. Relationship strength is how K3X answers that question, on every person and company record.
It is also one of the features we get asked about most, usually in the form “why does it say Leo knows this person better than I do?” This post explains exactly how it works.
The signals we count
Relationship strength is built from the emails and meetings K3X syncs from Google Workspace and Microsoft 365. For each pair of a teammate and a contact, we look at:
Two-way email. Threads where both sides have written. A reply counts for far more than a message that was never answered.
Meetings. Calendar events where both people were attendees and the meeting actually happened, rather than being declined or cancelled.
Recency. Every signal fades over time. An email from last week is worth roughly four times one from six months ago.
Consistency. Ten conversations spread over a year say more than ten emails in one afternoon.
We combine these into one of four levels: Very strong, Strong, Weak and None. The level on a company record is the strongest connection anyone on your team has with anyone at that company.
The signals we ignore
What we leave out matters as much as what we count. These do not move the score at all:
Newsletters, notifications and anything sent from a no-reply address.
Large group threads with more than 12 recipients, where nobody is really talking to anybody.
Sequence emails that got no reply. Sending a lot is not the same as knowing someone.
Calendar invites from people outside the meeting, such as an assistant booking on someone’s behalf.
The sequence rule caused the most debate internally. Early versions counted outbound volume, and our own sales team quickly topped every leaderboard by doing nothing more than running sequences. We took it out.
Why a word, not a number
We tried showing a score from 0 to 100 in the beta. People argued about the difference between 71 and 74, which means nothing, and stopped trusting the feature when the number moved by a few points overnight. Four levels are coarse on purpose. They change only when something real happens, like a first meeting or six months of silence.
Where the levels change
Level | Typical pattern |
|---|---|
Very strong | Regular two-way email and at least one meeting in the last 60 days |
Strong | Two-way email or a meeting in the last six months |
Weak | Some contact, but nothing two-way in the last six months |
None | No synced email or meeting at all |
These are typical patterns, not hard thresholds. A long run of meetings a year ago can still keep someone at Strong for a while.
Privacy
Relationship strength never shows the content of anyone’s email. It tells you that Priya has a very strong connection with a contact, not what they talked about. If a domain or a label is marked private in someone’s sync settings, it is left out of the calculation entirely, and admins can switch the feature off for the workspace.
Using it well
The teams that get the most out of relationship strength use it in three places:
Before outreach. Sort a list of target accounts by connection strength and start with the ones where someone already has a door open.
In pipeline reviews. A large deal with only weak connections is a risk, whatever stage it is in.
When people leave. When a rep moves on, their strongest relationships show up on the records they owned, so the handover starts with who to call.
If a level looks wrong, open the record and select it. K3X lists the emails and meetings behind it, newest first, so you can see exactly why it says what it says.
#inbox-sync
#relationships

Written by
K3X Team
K3X



