In twelve years on brokerage operations floors I never once had a shipper call me to end the relationship. Not once. That is not how accounts leave in freight.
They go quiet.
Twelve loads a week becomes eight. Eight becomes four. The tenders that do come start arriving on short notice, the ones nobody else wanted, the Friday afternoon pickup out of a market with no trucks. The Monday check-in call that used to happen every Monday goes ten days without anybody noticing. Your rep still likes the account. The account still likes your rep. Nothing has visibly broken.
Then one quarter later somebody runs a report and the account is down seventy percent, and by that point the decision was made a long time ago by somebody your rep has probably never met.
By the time it hits your numbers, they are already half gone.
Why the good version of this is invisible
The reason quiet churn is hard to see is that everything about a brokerage floor is built to look forward, not sideways. A rep's day is organized around what is on the board right now. Cover the loads that are on today. Quote the ones that came in this morning. Chase the POD from Thursday so billing stops asking.
Nothing in that day is organized around the load that did not come in. Absence has no ticket. There is no alert for the tender that was not tendered. If a customer sends you nothing on Tuesday, your system records exactly nothing, and nothing looks the same as a quiet week.
That is the whole problem in one sentence. Freight systems are very good at telling you what happened and nearly useless at telling you what stopped happening.
What the fade actually looks like
If you go back and study accounts that walked, the pattern is remarkably consistent. It is rarely price alone, which is the reason everybody guesses wrong about it. Price is the reason people give you afterwards because it is the polite one.
What actually shows up first, in roughly this order:
- Volume thins before it stops. The account does not go to zero. It goes to a third and stays there long enough that the new number starts to feel normal.
- The mix gets worse. You keep the hard lanes and lose the easy ones. That is the tell that somebody else got awarded the good freight and you are being kept around for coverage.
- Lead time shortens. You start getting tenders with less and less notice, because you are now the backup rather than the primary.
- The conversation gets transactional. Emails only, shorter, no small talk, and the questions are all about rate.
- The contact changes and nobody registers it. This is the one that kills accounts. Your relationship was with a person, that person moved to a different role, and the new one has no history with you and no reason to protect you.
Any one of those is noise. Three of them together, on the same account, inside six weeks, is an account that is leaving.
The uncomfortable part
Most brokerages I have seen do have this information. It is sitting in the TMS. Every tender, every load, every date, every contact, every lane. The data required to see a fade coming is almost always already in the building.
What is missing is anybody whose job it is to look at absence. And you cannot fix that by asking reps to look harder, because a rep with fifty accounts and a full board is not going to notice that account thirty-one has not tendered since the eleventh. That is not a discipline problem or a caring problem. It is an attention problem, and no amount of accountability meetings fixes an attention problem.
You are only ever as good as your last at-bat, because everybody else is knocking on your accounts' doors too.
What I would actually do about it
If I were running a floor tomorrow with no new software at all, I would do three things.
One, define quiet numerically, per account, and write it down. Not "we should keep an eye on them." An actual threshold. This account normally tenders nine to twelve a week. Below six for two consecutive weeks is a flag. The number is different for every account, which is exactly why nobody does it, and exactly why it works when you do.
Two, make somebody own absence. One person, once a week, whose entire job for that hour is to pull the accounts that are below their own baseline and hand them back to the rep with a name and a reason. Not a dashboard. A list of accounts and a sentence each about why they are on it.
Three, treat a contact change as a fire. When your person at an account moves, you have roughly one quarter of goodwill left and then you are a vendor again. That is the moment to get in front of the new person, and almost nobody does it because nothing in the system tells them it happened.
None of that is clever. It is just pointed at the right thing. Most retention effort in freight goes into being responsive when the customer calls, and the customer who is leaving has stopped calling.
Why I ended up building software about it
This is the problem I could not put down, and it is a large part of why FreightScout exists. Not because a brokerage needs another screen to stare at. Because the work of noticing an absence, across every account, every week, without getting tired or distracted or busy, is genuinely not human work. It is exactly the kind of thing you should hand to a machine so that your best rep gets the twenty minutes back and spends them on the phone call that saves the account.
The rep is still the one who saves it. That part does not change and should not. The relationship is the entire business. All I want is for nobody to find out too late.