Credentialing Was Late On Purpose

Ask anyone in healthcare staffing when credentialing should start and you'll get one word back. Earlier. Everyone says it. Almost nobody does it.

That's bothered me for a long time. When a whole industry agrees on a better way and then quietly doesn't do it, one of two things is going on. Either the better way isn't actually better, or there's a reason nobody says out loud.

I think it's the second one, and it took me a while to find it.

The excuses we've all used

The usual explanation is that credentialing teams are reactive. Buried, understaffed, always waiting on something. The work starts after the offer letter because nobody has the room to get ahead of it.

The other one has turned into folklore: candidates don't want to be pushed. A nurse who hasn't been placed yet isn't going to chase down references or upload a license. Ask too early and she stops answering.

Both of those are stories we tell. Neither one is the actual reason.

Part of it was math

Verification costs money per candidate. Not per placement. Per candidate.

Someone waits on hold with a license board. Someone calls a nurse manager for the fourth time about a reference that still hasn't come back. Someone assigns the skills checklist, builds the packet, and digs up seven years of employment history because that one facility asks for it.

Move all of that to submittal instead of placement and you've multiplied the cost by every candidate you submit who doesn't get the job. Most of them don't get the job.

So the sequencing wasn't sloppiness. Placement was the moment verification paid for itself, so verification waited for placement.

“

Credentialing started late because starting early meant paying to verify people you'd never bill for.

Which means "start earlier" was never really advice. It was a wish that kept losing to a budget.

But math doesn't explain the shadow credentialing team

Here's the problem with stopping there. If cost were the whole story, pre-offer credentialing work simply wouldn't happen. It happens constantly.

At most agencies, someone is already collecting references, pulling licenses, and chasing checklists long before an offer exists. That someone just isn't in credentialing. It's the recruiter.

Nobody decided this. No leadership team sat down and assigned verification work to the sales side of the building. It settled there on its own, and once you see why, the offer letter stops looking like a cost threshold.

The offer letter is an ownership line

Before the offer, the recruiter is the only person in the building with a personal stake in this specific candidate clearing. Their number depends on it. So the work drifts to them, because that's where the caring is.

After the offer, the file is real enough that a salaried team with no stake in this particular placement can justify picking it up.

That's the line. Not cost. Ownership.

And it worked, mostly, which is why it lasted so long. But look closely at what the recruiter is actually incentivized to produce. They want a packet complete enough to submit. Credentialing wants a packet that holds up when a facility audits it eighteen months from now. Those two goals overlap most of the time. They come apart on exactly the cases that end up costing you money.

Credentialing never argued for later

Try the counterfactual. Give a credentialing team unlimited capacity and ask them when they'd want to start. Every one of them says first contact. Start collecting at the top of the funnel, analyze it while there's still time to fix problems, know before you submit whether this person can actually clear.

They were never the ones asking for later. "Start earlier" wasn't a debate credentialing lost. It was a request the org couldn't fund, so the org routed around it and hoped the recruiter would be thorough.

We were rationing the wrong thing

Waiting protects credentialing hours. It spends the candidate.

In travel, per diem, and locums, the candidate is the scarce part. A clinician who can work this week doesn't stay available for long. And while her file sits there waiting on a signed offer before anyone verifies anything, she's on the phone with two other agencies.

So we protected the thing we had plenty of and spent the thing we didn't. For years.

What actually changed

Cost is the boring half. Yes, agents make a license check, a reference chase, and a document read cost close to nothing per candidate, so the arithmetic that made waiting correct stops working. You can verify everyone you submit now, because the ones who never place don't sink the budget.

The more interesting half is the incentive problem, because it just disappears.

“

An agent has no quota, no commission, and no opinion about which candidate places. You stop needing anyone to care.

That's the part I didn't expect. The old arrangement needed a human whose bonus was attached to this file, which is precisely what made the work fast instead of correct. Move it to an agent and rigor stops depending on whose paycheck is downstream of the outcome. You're not moving work earlier and hoping the person doing it is thorough. You're removing the requirement that anyone be thorough.

The candidate folklore falls apart too. Nurses don't resist credentialing. They resist a portal, a password, an eight-page PDF, and four different people asking for the same document. Ask well and ask once and they'll respond at submittal, before there's any offer on the table.

And then submittal stops being a handoff and starts being a specification. Right now a submittal says: here's a candidate, go. The requirements surface later, facility by facility, usually the moment they bite. Flip it and the submittal carries the facility's requirement set with it, and the file gets worked against a known target from the first minute.

The submittal stops being a bet you have to win before the work starts. It becomes the work, already running.

Where I'm still not sure

Verifying candidates who never place still costs you something even when it's cheap. Candidate goodwill. Source-side rate limits. Your own review queue. There's a right level of speculative verification and I don't know where it sits yet. Anyone quoting you a precise number is selling you one.

The direction I'm confident about. And I'd stop calling credentialing teams behind for starting late. They weren't behind, and they weren't wrong. They were correct under conditions that don't hold anymore.