A pattern independent practice owners describe repeatedly: they outsourced billing to fix one problem, and ninety days in, they are quietly running parallel in-house coverage because the new vendor is not delivering. One retrospective line shows up in thread after thread practitioners open about billing regret: "should have been the other way around." Meaning: start in-house, outsource only when growth forces it.
The first ninety days with a new billing company either validate the switch or expose the pattern that made it a mistake. Five failure patterns show up with enough corpus frequency that each one is worth diagnosing before you commit further, or pull out.
What "working" looks like: four numbers, not ten
Before diagnosing what is going wrong, fix the baseline. Practitioners who run clean revenue-cycle reviews track four numbers, and every failure pattern below ties back to at least one of them.
- Days in AR: under 40 for commercial payers, under 50 overall. Creeping past these means claims are aging faster than they are resolving. MGMA's revenue-cycle benchmark guidance tracks days in AR as the earliest leading indicator of rev-cycle drift.
- Clean claim rate: 95 percent or higher. The share of claims accepted on first submission without rework.
- Denial rate: under 8 to 10 percent, broken out by payer. A single payer driving the composite number up is a different problem than a system-wide creep. AAPC industry data places the national average closer to 15 to 20 percent and the acceptable threshold near 7, so anything above 10 in your post-transition book is a meaningful signal.
- Net collection rate: dollars collected divided by dollars collectible after contractual adjustments. Any gap between gross and net is leakage, and leakage has causes.
If your billing company cannot produce these four numbers on demand, that is already one of the patterns below. Skip to Pattern 3.
The five failure patterns
Pattern 1: Onboarding drag
Most billing contracts promise full operational handoff in 30 to 45 days. Practices that actually track the curve see 60 to 90 days before the new vendor is clearing claims at parity with whatever preceded it. During that gap, claim submission stalls, eligibility verifications lag, and days in AR climbs before any other number flinches.
The symptom is specific. Days in AR jump 10 to 15 points over month one, then plateau rather than recover in month two. Clean claim rate also dips during the same window because the new team is still learning your payer mix and coding quirks.
Onboarding drag is structural, not a failure signal on its own. It becomes one when the curve has not corrected by day 75.
Pattern 2: Specialty-mismatch undercoding
A biller without specific specialty experience will consistently undercode. For anesthesia, behavioral health, and other code-dense specialties, the corpus estimate lands at 5 to 10 percent of collections lost to conservative coding choices that would have been billable at a higher level with the right documentation pairing.
The number that surfaces this is net collection rate. Volume of claims looks healthy, denial rate looks fine, AR days look fine, but collected dollars run below what the same case mix produced under the previous setup. That gap is the undercoding tax.
Specialty-mismatch undercoding does not resolve on its own. If the vendor is not specialty-staffed by day 45, the pattern is permanent until you switch.
Pattern 3: Black-box reporting
The non-transparent biller pattern is familiar. Monthly reports arrive with top-line numbers but no drill-down. You cannot see denial reasons by payer, claim status by age bucket, or which providers are carrying the AR load. When you ask, the response is a delay, then a narrative summary, then another delay.
This is the one pattern that disqualifies a billing company before the 90-day mark. The four baseline numbers require source-level detail to be useful. A vendor that will not share that detail is not a billing partner, it is an intermediary you are paying for access to your own data. If dedicated-contact access and raw report exports are not in the contract, diagnose this one on day one.
Pattern 4: Inherited-mess masking
The subtle pattern. Your old biller left behind aging claims, unsubmitted rework, unresolved appeals, and credentialing gaps. The new biller inherits that backlog alongside your new claim volume. Three months in, AR days look bad, collections look bad, and the instinct is to blame the new team.
The diagnostic fix is an AR audit at transition. Segregate pre-transition aging from post-transition activity. Track the four baseline numbers against the post-transition book only for the first 120 days. If post-transition AR aging is clean and the inherited backlog is dragging the composite number, that is not a new-biller failure. Without this split, practices fire a competent vendor for a predecessor's mess.
Pattern 5: Fee-structure misalignment at scale
Percentage-of-collections fee structures work at a specific collection volume and stop working above it. The typical range is 4 to 10 percent of collections, depending on specialty and claim complexity. As a practice scales, the absolute dollar amount of that percentage outpaces the operational cost of running billing internally, often by month nine or twelve.
The pattern shows up in performance that looks fine on all four baseline numbers while the bill from the vendor grows faster than collections themselves. This is the "should have been the other way around" realization: vendor economics that justify outsourcing at $1.2M in collections do not survive at $2.5M. For the economics at your actual volume, see the pillar on outsource vs. in-house billing.
When to switch vs. give it more time
Ninety days is a useful diagnostic window but a bad automatic trigger. Three rules for deciding:
- Pattern 3 is a switch signal by day 30. Black-box reporting does not resolve with time. If the vendor is not producing payer-level denial breakdowns and AR aging by bucket at day 30, they will not be producing them at day 120.
- Pattern 1 gets until day 90. If days in AR have not trended back toward baseline by the end of month three, the vendor is not going to catch up. Switch.
- Patterns 2, 4, and 5 need a clean diagnosis first. Undercoding, inherited mess, and fee misalignment can all look like failure without being one. Run the pattern's specific test first: specialty coding review, pre-versus-post AR segregation, or a cost comparison against in-house at your current volume.
The worst switch is the one made on composite numbers without diagnosing which of the five patterns is actually driving them. That sequence is how the "switched back, then switched again" subset of practice owners describe their billing history.
The diagnostic habit
The four revenue-cycle numbers, run against the post-transition book, surface which of the five patterns is running. From there the decision is whether the pattern resolves with time, with configuration changes, or only with a switch. Most practices that switch back a second time did not run this diagnostic the first time. The pillar on outsource vs. in-house billing covers the economics of the structural decision, and reducing claim denials covers the operational inputs to Pattern 2 and Pattern 3. Both run off the same four numbers.
Frequently asked questions
How long should I give a new billing company before deciding it's not working?
Ninety days is the working window. Days in AR should trend back toward under 50 overall by month three, clean claim rate should hold at 95 percent or higher, and denial rate should stay under 10 percent for the post-transition book. If any of these three have not corrected by day 90, the vendor is not going to correct them in month four.
What is a normal denial rate for outsourced billing?
Under 8 to 10 percent is the working baseline, broken out by payer. A single payer pushing the composite number up is a different problem than a system-wide denial creep. A vendor that cannot produce the payer-level breakdown is Pattern 3, not a denial-rate conversation.
Can I switch billing companies mid-year without losing AR?
Yes, with a pre-transition AR audit. Segregate aging claims from new claims, document clearinghouse credentials, and keep the outgoing vendor on a 45 to 60 day wind-down for residual claim submissions. Most AR loss at transition traces to skipped audit steps, not transition itself.
How do I know if my biller is actually undercoding?
Pull a sample of 25 to 30 claims from the last 60 days and route them through a specialty-specific coding review, either internal or from a second biller with the specialty background. If the review flags 10 percent or more of claims as under-documented for higher-level billing, Pattern 2 is running.