Cost of Building Insurance Operations In-House: Build vs Buy

When a founder says “we need to figure out billing,” they are describing five jobs. Each has its own talent pool, its own failure modes, and its own timeline. Treating them as one job is the most common reason in-house insurance operations underperform.

1. Health plan contracting

Getting a contract with a payer means a negotiated agreement between a legal entity and the plan: rates by code, terms, credentialing requirements, timely filing windows, and renewal cycles. It is part legal, part sales, part data. A good contracting lead knows which payers in which states cover the lives you want, what the rate ladder looks like, and when a “we’re not accepting new providers in your specialty” is real versus an opening position. The published side of that rate ladder is on reimbursement rates by state and CPT code.

Timeline for a competent internal hire to reach national density: one to three years. Talent: scarce, senior, expensive. Most people with this background did it for one health system in one region.

2. Credentialing and enrollment

Once the entity has a contract, each clinician has to be credentialed (primary source verification of license, education, board status, malpractice history) and enrolled (added to the payer’s roster under that contract). Credentialing is document-heavy and combinatorial. Ten clinicians, eight states, four payers is 320 enrollments, each with its own portal, form, and follow-up cadence. The difference between a contract and a roster is covered in what in-network nationally means.

Timeline: 90 to 120 days per clinician per payer on average, longer when applications bounce. Talent: credentialing specialists, roughly one per 40 to 50 clinicians, plus someone who knows how to automate the repetitive parts.

3. Eligibility and benefits verification

Before a visit, someone has to confirm the patient’s plan is active, the service is covered, the clinician is in that plan’s network in that state, and what the patient owes. Done manually, this is a phone call and a portal check. Done properly, it is an API call at the moment of booking that returns a yes/no and a cost estimate.

Getting this wrong is expensive twice: eligibility errors drive a meaningful share of denials, and a clunky insurance step drives a meaningful share of funnel drop-off. Talent: this is engineering plus payer-data knowledge, not billing staff.

4. Coding and claims submission

Every visit becomes a claim: CPT codes, modifiers, place of service, diagnosis codes, rendering and billing provider, submitted through a clearinghouse to the payer. New billing operations should expect a rough first six months while workflows stabilize. When claim volume outpaces submission capacity, cash slips a week for every week of backlog.

Talent: certified coders and billers, plus EHR-to-clearinghouse plumbing.

5. Denials and accounts receivable

Some claims come back denied. Denials cost new programs a meaningful share of revenue while the team learns each payer’s patterns. The gap is lost revenue and staff time on appeals. This work is pattern recognition by payer and discipline about timely filing.

Talent: AR specialists with payer-specific experience.

Which ones to build

Here is the honest map.

Buy claims and denials until the volume justifies a team, and telehealth revenue cycle management covers coding, claim submission, and denial management.

The pattern: outsource the labor, own the asset. Most vendors get you the first half. Few are set up to hand you the second. Owning the asset starts with the entity: compare rent a PC vs own a PC, and see how your own PC is papered as an MSO-PC structure for telehealth. If you launched on a rental vendor, read the Bridge insurance billing alternative.

See how the five workstreams run on the platform, and how you own your payer contracts. If you are on a vendor now, that move is switching from a network rental vendor.

Next step

Model the build cost for your company with the build vs partner calculator.