When cash drops, most people chase theories in whatever order the panic suggests. Somebody blames a payer. Somebody blames the season. Somebody wonders about the billing company. There’s a better way to run this: rule things out, in order, starting with what it can’t be. We’ve run this sequence live on calls, and by feel it takes ninety minutes. With the checks below, it takes closer to ten.
Before step one, do one small thing: write down the date the drop started, as best you can tell from the deposits. Every check below needs a window, and a fuzzy window makes every answer fuzzy.
Why did my practice’s cash suddenly drop?
One of four causes, checked in order: a recording backlog (rule it out first), claims that never left, rulings nobody worked, or a payer that changed its behavior. The checks below run the whole sequence in ten minutes.
Step 1: It isn’t recording
This one goes first because it’s everybody’s favorite theory and it dies in one sentence. Cash lands in the bank whether or not payments get recorded in your billing system. So if actual deposits fell, paperwork lag didn’t do it. The check: put this quarter’s weekly bank deposits next to last quarter’s. If the deposits themselves are down, recording is ruled out. Move on. (If deposits are fine and only your reports look scary, you have a recording backlog, not a cash problem, and that’s a different, easier day.)
Step 2: Did the claims actually leave?
Billed is a status your software sets. Sent is an event in the world. The check: sample twenty claims from the weeks where cash fell and look for two proofs on each: accepted by the clearinghouse, the postal service that carries claims to insurers, and acknowledged by the insurer. If a chunk of your sample has no acknowledgment, stop here. Your leak is upstream, claims dying between your software and the insurer’s front door, and no amount of payer-blaming will fix it.
Step 3: Are the rulings being worked?
If the claims arrived, the next question is what happened when answers came back. The check has two halves. Pull the denials from your drop window and split them one way: touched in the last 30 days, or not. Then find your rejections, the claims that bounced before the insurer ever processed them, because those live in a different queue and often in a report nobody reads. A large untouched pile in either place is your leak, and every one of those claims has a filing clock running.
Step 4: Did a payer change on you?
If the claims left and the answers are being worked, now, and only now, look at the payers. There are three ways one breaks pattern. Slower: their days-to-pay drifts outside their own twelve-month pace. Lighter: the checks arrive on time but the cents on the dollar drop, sometimes to zero. Quieter: fewer payments arrive at all, even though every payment that does arrive looks normal. The check: take your top five insurers by dollars and compare each one’s current month to its own last year, on all three questions. Any insurer failing any question is your answer, with a name and a dollar figure attached. The full version of that comparison is its own story.
Why the order matters
Each step is cheap to check and rules out everything the later steps would waste time arguing about. Every real cash drop we’ve traced landed in step two, three, or four. Almost every argument that preceded the tracing was about step one.
Here’s what a live run looks like. One practice, deposits confirmed down against last quarter, so step one closed in a minute. A twenty-claim sample all carried insurer acknowledgments, so submission cleared. The denial pile was current, touched inside thirty days, so step three cleared too. Step four found it: one insurer’s payment count had thinned for six weeks while every payment that did arrive looked perfectly normal. The quiet break, invisible to every speed metric, obvious the moment someone asked how much money should have arrived. The sequence didn’t make anyone smarter. It just stopped the argument early enough to leave time for the answer.
When to stop doing this by hand
Keep a log as you go: the drop window, each step, what the check showed, ruled out or found. One page. It turns the drill into a record, which matters twice. It keeps the room from re-arguing step one on Thursday, and if the drill ever runs again, the last log is the head start.
The sequence is a fire drill. It works, and you shouldn’t need it. The permanent version is smoke detectors: the same four checks, running on their own schedule, flagging a break the week it happens instead of the quarter after, the same rhythm that makes month-end a decision meeting instead of an autopsy. If you’ve run this sequence more than once, that’s the signal to stop drilling and start detecting.
We’ll run the whole sequence on your data and show you which step your cash drop lives in. Grab 30 minutes with us. Prep nothing.