Cost to Collect

What cost to collect is

Cost to collect measures how much the organization spends on revenue-cycle activities for every dollar it actually collects. The core inputs are your total RCM-related expenses over a period and the cash posted or net patient service revenue for that same period.

RCM expenses include billing and collections staff, vendor fees, statement and lockbox costs, clearinghouse and practice management system fees, and sometimes related IT and overhead. The more labor and friction you have in eligibility, authorizations, claim edits, and follow up, the higher this number climbs.

Most leaders look at cost to collect as a ratio or percentage. For example, if you spend $40,000 to generate $1,000,000 in cash, your cost to collect is 4 percent.

Why cost to collect matters operationally

Cost to collect translates RCM efficiency into dollars. A few percentage points in either direction can mean hiring or layoffs, funding a new clinical program, or falling short of payroll during a census dip.

For operators, the metric exposes where you are burning money:

• High manual touches on eligibility and authorizations
• Excess rework from denials and avoidable re-bills
• Expensive vendor contracts that do not move collections
• Overstaffing in one function paired with bottlenecks in another

Because cost to collect ties directly to expense, changes in workflows, automation, and staffing models should show up in this metric within one or two quarters. If claim quality or denial rate improves but cost to collect does not move, you may simply have shifted work without actually reducing effort.

How cost to collect is used and read

Leaders typically track cost to collect over time, by payer class, and sometimes by program line. You want a consistent methodology so month-over-month and year-over-year trends are meaningful.

When the ratio increases, you ask: did collections drop, did expenses rise, or both. A temporary increase can be acceptable if you are investing in new staff or tools to fix chronic denials. A persistent increase is a red flag that you are doing too much manual work or paying too much for the dollars you bring in.

It is also common to compare cost to collect by payer: commercial vs Medicaid vs self-pay. Low reimbursement behavioral-health contracts will always have a higher cost to collect percentage, because the same work brings in fewer dollars. That is not always a process failure, but you should know where you are underwater so you can price services, renegotiate, or limit exposure.

Beyond the main formula, a few variations are worth knowing. Each answers a slightly different question.

Imagine a behavioral-health group for a calendar quarter.

Total RCM-related expenses for the quarter:

• Billing team salaries and benefits: $75,000
• Authorization and utilization review staff: $45,000
• Clearinghouse and PMS fees: $10,000
• Statement printing and payment portal fees: $5,000 Total RCM expense = $135,000.

Cash collected for services in the same quarter from payers and patients totals $2,250,000.

Cash-based cost to collect = $135,000 ÷ $2,250,000 = 0.06, or 6 percent.

Now you separate a high-friction carve-out payer. You estimate $40,000 of your RCM expense is tied to that payer and you collected $300,000 from it.

Payer-specific cost to collect for the carve-out = $40,000 ÷ $300,000 = 0.1333, or about 13.3 percent. That single payer is more than double the organization-wide cost to collect. That insight supports renegotiating the contract, tightening authorization processes, or capping volume from that payer.

• Leaving out authorization and utilization review staff from RCM expense, so the cost to collect looks artificially low even though clinicians are constantly pulled into concurrent review calls.
• Mixing time periods, such as using a quarter of RCM expenses but dividing by six months of collections because of a delayed cash posting cleanup, which produces a meaningless ratio.
• Not backing out non-patient-service cash like grants, donations, or pharmacy rebates, which can make the metric look better than it is for the actual treatment programs.
• Comparing payer-specific cost to collect without allocating expenses properly, for example blaming Medicaid for a high cost to collect when most of the manual follow up is actually for a single commercial carve-out.
• Treating a short-term spike, such as a one-time EHR conversion or cleanup project, as a permanent trend and cutting staff too aggressively, which then drives up denials and rework later.

Behavioral health has structurally higher cost to collect than many medical specialties because so much work happens before and during the episode. Benefit checks, carve-out identification, and repeated concurrent authorization updates add labor without increasing reimbursement. If you do not count those costs inside RCM, you end up underestimating what it really takes to bring in a dollar.

Manual eligibility and authorization tasks are especially heavy in residential, PHP, and IOP programs. Staff reverify benefits when a client steps up or down levels of care, chase new auths when days are exhausted, and fax clinical notes to UM nurses. This is often done by clinical or admissions staff, not just the billing team, so the true cost to collect is split across departments.

Carve-outs create another cost driver. A client might have medical benefits with one payer and behavioral health with a separate vendor. That means double benefit checks, different portals, and more frequent denials for "no auth" or "not in network." Each of those touches increases cost without guaranteeing payment.

Behavioral health programs that bring cost to collect down usually do it by automating repetitive verification, pre-billing checks, and status follow up. The work does not disappear; it becomes concentrated on exceptions and clinical issues where humans add real value, rather than on checking yet another portal to confirm the same benefit details.

For cost to collect, AI is most useful at reducing the manual touches that drive the numerator of the metric. Agents can handle high-volume, rule-based tasks like checking behavioral-health benefits on multiple portals, pulling authorization requirements, queuing up concurrent review deadlines, and checking claim status without a human logging into six different systems.

Supabill runs this through specific agents: a benefits-verification agent that remembers payer-specific carve-out rules, a claims-scrubbing agent that applies payer edits and authorization rules before anything goes out the door, and a denials agent that reads every 835, classifies CARC and RARC codes, and routes workqueues by root cause. Humans still own how expenses are categorized, which investments are worth making, what to negotiate with payers, and when higher cost to collect is acceptable to maintain access or support a strategic program.

What expenses should be included in cost to collect?

At minimum, include all direct revenue-cycle expenses: billing and collections staff salaries and benefits, authorization and utilization review staff if separate, vendor fees for clearinghouses, statement vendors, payment portals, and revenue-cycle related software. Many organizations also include allocated overhead such as a portion of IT, HR, and facility costs tied to the RCM function. The key is to define the scope clearly and use it consistently over time so trends are real, not artifacts from changing what you count. HFMA's MAP Keys provide common definitions organizations adapt for their own reporting. Source

Should cost to collect be based on cash or net patient service revenue?

Either can work, as long as you are consistent and understand the tradeoffs. A cash-based approach uses total cash posted in the period, which reflects the actual dollars your RCM work produced but can be distorted by large timing shifts or cleanups. An approach based on net patient service revenue uses accrual accounting and lines the metric up with financial statements, but it can hide cash-flow issues if your AR is growing. HFMA and MGMA both describe cost-to-collect definitions that allow for either method. Source

How do concurrent authorizations and utilization review affect cost to collect in behavioral health?

Concurrent auth and utilization review add labor in the middle of the episode: staff monitor authorized days, track review dates, prepare clinical packets, and call payers for continued-stay decisions. In residential and PHP/IOP, those steps repeat every few days or weekly across many clients. If you pull clinicians or case managers into that work, you are effectively shifting RCM cost into program budgets. To measure the real cost to collect, include the time and salary expense for those tasks, even if they sit outside the billing department.

Can a higher cost to collect ever be acceptable?

Yes. A temporary increase can be reasonable if you are investing in new staff, systems, or cleanup work that should reduce denials and bad debt over time. It can also make sense to accept a somewhat higher cost to collect for strategic programs, such as high-acuity adolescent residential, where access and outcomes matter more than immediate efficiency and where payers are difficult to work with. The key is to pair cost to collect with quality metrics, denial trends, and payer-level performance so you can see whether the extra spend is buying better collections, better stability, or better care.

How should I use cost to collect in payer negotiations?

Calculate a payer-specific cost to collect by allocating staff time, vendor fees, and write-off patterns to each major payer or carve-out. If one plan requires far more manual eligibility checks, concurrent auths, and appeals for the same or lower reimbursement, you can use that data in rate and contract discussions. Show the payer that your administrative burden is higher for them and tie that to denial patterns and underpaid claims. While cost to collect is not a published benchmark in most contracts, it is credible context alongside denial rate and first-pass resolution when you press for rate increases or process changes. Source

• HFMA — MAP Keys