Your CRM Knows Who Paid. It Doesn’t Know Who Matters.
How we’re teaching our software to remember the relationship history that usually disappears after the meeting ends.
We all know business runs on relationships.
But recently, at our firm, we realized the software we use to run the business knows almost nothing about them.
Our CRM knows who got billed.
Who paid.
Who signed the proposal.
What stage the opportunity is in.
How much the lead is worth.
Ours even has a field where we can mark that we don’t like someone. The “client credit score.”
But it does not know how much that person matters.
It does not know that the contractor on this new job is the same one we fought with last time.
It does not know that the architect is polite over email but keeps steering the owner away from our recommendations.
It does not know that the developer bringing us work also keeps bringing us the jobs nobody wants to touch.
The firm knows all of that.
It just lives in hidden places that have been previously impractical to capture.
In a meeting comment.
In a project huddle.
In the five-minute vent after a frustrating review.
And then it disappears right before you need it.
So we’re trying to study ways we can capture this data once and for all.
The person who matters is not always the person who pays
It’s dangerous to treat “the customer” like the center of the universe.
Design and construction do not work that cleanly.
The owner may pay the invoice.
The architect may control the room.
The contractor may control the pace.
The developer may control the next three jobs.
The owner’s rep may be the person who quietly makes everything easier, or quietly makes everything harder.
So we stopped treating everything like one giant client record. We tried that, and it wasn’t giving us good predictive signals.
A billing customer pays us.
A company is the real-world business behind the work (multiple per project).
A person is a person (multiple per company per project).
The person carries reputation, warmth, and history.
We stopped obsessing over billing customer (top down) and started studying the people (bottom up).
The point of contact who loved you leaves, and the proposals leave with them.
The person who drove your team insane is suddenly signing contracts for your biggest account.
An account can look healthy right up until one project quietly goes nuclear.
Your CRM doesn’t store all that information. That is unless, you want to make your job managing CRM instead of closing deals.
We wired up the memory
Think about how much relationship data your team creates in a normal day.
Meetings.
Calls.
Emails.
Internal threads.
Site visits.
“Be careful, we’ve worked with them before.”
Nobody is documenting all of that by hand no matter how much HR tells them to.
So we wired it up.
Meeting transcriptions.
Call recordings.
Internal discussions.
Emails.
The system reads the stream and starts connecting the dots.
Who works for whom.
Who brought in what.
Who keeps showing up around risk.
Who makes projects smoother.
Who makes them harder.
All of it goes into a living relationship graph.
After only two months, it’s grown to tens of thousands of connections deep.
Built from conversations that used to evaporate.
The lesson so far is:
Just wire it up.
The company starts having a memory it never had before. Once you have the data, you can figure out how to use it later.
In some ways, it’s only possible to know what it can do for you until you capture it first.
So… just wire it up.
What we found
Things got interesting when we looked at the data.
Our graph had flagged a certain outside firm as a scope-creep risk.
Not once.
Four times.
Same name.
Same pattern.
Tense communication.
Extra internal effort.
Coordination that got harder than it should have.
The graph knew.
We did not act on it.
On paper, things looked fine.
In the room, people braced when the name came up.
The gut said run.
The paperwork said fine.
The difference between your collective gut instinct and the paperwork is reflected in your margins.
A real example: getting fired
A few weeks ago, a person showed up on one of our jobs.
We do not have a great relationship with them.
They had no contract with us.
This week, we got fired from that job.
Their opinion of our design was the reason.
For the record, their objections were wrong and not code-compliant.
Their recommendations to the client were illegal.
Didn’t matter.
They controlled the narrative.
Being right did not save us.
So we looked at the graph.
That one person had more than twenty relationships with other clients and projects in our world.
We were not looking at a difficult person.
We were looking at blast radius.
One confidently incorrect person with that much reach can do damage way beyond the project they are standing in.
Blast radius is the cost of ignoring influence
Once relationships live in a graph, influence stops being a gut feeling.
You can walk the connections.
One step out.
Two steps.
Three.
How many people, projects, companies, and clients can this person touch?
A difficult person on one job is annoying. A difficult person tied to twenty other relationships is a business problem.
And if they like you, great. That is a huge asset.
If they do not, pretending they do not matter is stupid.
And so now we’re finally getting a chance to measure the true cost of having an ego.
The old answer would have been easy.
“They are wrong.”
“They do not know what they are talking about.”
“They are recommending an illegal installation.”
Wouldn’t it be nice if being right mattered in business?
If it did, I could have retired at fourteen. Just ask my dad.
If a wrong person has twenty-plus relationships in our world, being right is not enough.
Ignoring them because your pride is hurt is not principled.
It is expensive.
The trick is to get the signal as quickly as possible that a high blast radius risk just showed up at the OAC meeting.
Where we are now
At PermitZIP, we already have passive capture running across meetings, calls, and emails.
We already extract people, companies, projects, and relationships into a living graph.
We already have tens of thousands of connections.
We already automatically score billing customers on credit and relationship health.
We already track warmth and reputation signals on people.
The next phase is connecting the graph directly into scoring and risk.
Who keeps showing up around problems?
Who has reach?
Who has low trust?
Which projects are exposed?
Which relationships need attention before they become expensive?
I want to know when someone pays well but hires contractors that burn the team.
I want to know when someone is difficult but influential regardless of our contractural relationship.
I want to know when the client is not the issue, but the third party around them is.
I want to know when a person with a bad history just showed up somewhere they can do real damage.
And I want to know when I need to swallow my pride because the relationship has more reach than my opinion of the person.
We’re rolling out a system upgrade this week. If it works, I’ll write more and start to release the technical findings in the white paper section of the site.
Stay tuned.