Do You Still Hire Payment Posters on 10-Key Speed? Posting Isn't the Job Anymore - Reading the 835 Is.

Do you still test payment posters on 10-key speed? Most billing companies do. It is on the interview sheet, right under "two years' experience in athenaOne or eClinicalWorks," and it made perfect sense when a poster's day was keying paper explanation-of-benefits statements line by line. It makes almost no sense now, and the gap between what you screen for and what the job has become is quietly costing you your best people.

Here is the pressure behind that. US small-business adoption of AI reached 66% in 2026, up from 55% a year earlier. But roughly 70% of owners say their staff need more training to use it well. The tools stopped being the hard part. The skills became the hard part — and in revenue cycle work, the skill that matters changed shape without anyone updating the job description.

Which seats in a billing company actually change

Not all of them, and not equally. Walk your own org chart:

• Payment posting. Electronic remittance advice already auto-posts the clean lines. What is left is the exceptions: takebacks, offsets, secondary crossovers that did not cross, credit balances, and remits where the payer's adjustment does not match your contracted rate. That is not data entry. That is variance analysis.
• Charge entry. Shrinking fastest where the client's chart system pushes charges through cleanly, and surviving where superbills still arrive by fax or scan from small specialty offices.
• AR follow-up. The biggest change. The specialist stops being the person who assembles the appeal and becomes the person who judges whether the assembled appeal is right — more decisions per hour, fewer keystrokes.
• Coding. Barely shrinking. A CPC or CCS is still the one who decides whether the documentation supports the modifier, and now they are reviewing more suggested codes per day than they used to assign.
• Credentialing. Almost unchanged in judgment, much faster in legwork. Somebody still owns the CAQH re-attestation clock every 120 days and the payer enrollment paperwork; the chasing and the form-filling around it compresses.
• Client account management. Grows. When the mechanical work compresses, what clients pay for is someone who can explain why their net collection rate moved.

The job did not disappear; the unit of work moved from the keystroke to the decision. That single sentence should rewrite your interview sheet.

What a new hire needs to be taught in week one

Forget the two-week software tour. A person who will be reviewing machine-assembled work needs to be able to spot a wrong answer by Friday, and that is a teachable, testable skill. Here is a week-one curriculum that works in a 20-person billing shop:

Day one: read twenty real 835 remittances. Not a slide deck about remittances — actual files from actual payers, with the client names masked. Find the allowed amount, the contractual adjustment, the patient responsibility, and the reason codes. Say out loud what happened to each line.

Day two: tell a rejection from a denial. A claim bounced at the clearinghouse and acknowledged back on a 277CA never reached the payer. A claim denied on the remit did. New hires confuse these constantly, and an agent that confuses them will happily "appeal" something that was never adjudicated. Teach the difference with real examples in front of them.

Day three: work a machine-built worklist and hunt for the error. Hand them a batch where a lead has deliberately seeded two wrong calls — a claim past its filing window, and a denial where the cited policy is the wrong year. If they do not find both, they run another batch.

Day four: find the source. Given a denial, locate the payer's own written policy for that date of service — the coverage determination from the Medicare contractor, or the commercial payer's medical policy — and cite it. This is the single highest-value habit in the building: never argue from memory, always argue from the policy in force on the date of service.

Day five: sign something small. Under a lead's review, they put their name on three low-dollar appeals. Ownership from the first week is what makes the review habit stick.

flowchart TD
  A["Day 1: read 20 real 835 remittances"] --> B["Day 2: sort 277CA rejections from payer denials"]
  B --> C["Day 3: work an agent-built worklist"]
  C --> D{"Did they catch both seeded errors?"}
  D -->|No| E["Lead walks the CARC and the policy line by line"]
  E --> C
  D -->|Yes| F["Day 5: signs three low-dollar appeals under review"]
  F --> G["Week 3: writes worklist goals for their own client"]

What stops being a hiring requirement

Drop the 10-key test. Drop the typing speed minimum. Drop "must have three years in AdvancedMD" as a hard filter — system-specific muscle memory is now maybe two weeks of ramp, not three years. And stop treating memorized modifier lists as a proxy for competence; the person who knows where to look up modifier 25 versus 59 against the payer's current edit rules will beat the person who memorized last year's list.

Add three screens instead. First, give the candidate a remittance and ask them to explain it in plain English, the way they would to a practice manager. Second, give them a short machine-written appeal cover letter with one factual error in it and see if they catch it. Third, ask them to describe a time they were told something confidently and checked it anyway. That third one predicts more about review work than any certification.

The certifications still worth paying for

Coding credentials — the AAPC's CPC and CPB, AHIMA's CCS and RHIT — did not lose value. They gained it, because they certify judgment about documentation and code selection, which is exactly the part that stays human. Keep paying for exam fees and study time.

What is missing from every one of those curricula, as of mid-2026, is review skill: how to audit work you did not do. Nobody sells you that as a certificate. You build it internally, with seeded-error batches and a monthly sample audit where a lead re-works ten closed accounts from scratch and compares. Put it on the calendar or it will not happen.

Twelve months of growth without three more desks

Illustration figures — substitute your own. Assume your company collects $28M a year for clients on a 6% fee, that your book will grow 35% next year, and that your historical staffing ratio would require one more payment poster and two more AR follow-up specialists to absorb it. Assume a loaded cost of $52,000 for a poster and $58,000 for an AR specialist. Assume instead you hire one senior denials analyst at $72,000 and put twelve existing staff through a real training program.

About $82,000 of avoided cost in year one, and — the part that matters more — a team that can absorb the next 35% without another hiring cycle. Prove it by tracking accounts touched per specialist hour and overturn dollars per specialist hour, monthly, by name. If those two numbers do not move within a quarter, your training was a slide deck, not training.

Where the people problem does not go away

Two honest limits. First, this restructuring is genuinely harder on the person who has posted payments for eleven years and is fast and accurate and proud of it. Some of those employees make the jump to variance and credit-balance work beautifully. Some do not want to, and pretending otherwise is how you lose them in month four. Have the conversation early, in person, with a named path and a raise attached to it.

Second, review fatigue is real. A specialist who approves 391 packets a month will eventually stop reading closely — that is human, not a character flaw. Cap the review load, rotate people through different denial categories, and keep the seeded-error batches running quarterly so the habit stays alive.

Frequently asked questions

Do I lay off my payment posters?

Most owners do not need to. Growth absorbs the change if you plan for it — you stop backfilling posting seats and move the people you have into exception and credit-balance work, which pays better and is harder to outsource.

How long before a new hire is actually productive?

Faster than before on the mechanics, slower on judgment. Expect them working a real worklist within a week and trustworthy on unsupervised appeals somewhere between six and twelve weeks, depending on how many seeded-error batches you put them through.

What do I tell clients about who is doing their work?

The truth, in writing, before they ask. Say which steps are assisted, who signs, and how you audit. Practice administrators are far less nervous about assisted work than about finding out later that they were not told.

Should I hire remote for these seats?

Review work travels well, but week-one training does not. If you hire remote, get the first week in person or on camera all day, with a lead reading remittances alongside them. Skipping that is why remote AR hires wash out.

CallSphere builds AI voice and chat agents that answer business phone lines and web chat, book appointments, and capture leads around the clock. In a billing company, the honest fit is the phone: patient statement calls and practice inquiries that pull trained specialists out of the worklist all day. Handling the routine ones consistently is what gives the people you just retrained the hours to do the judgment work you hired them for.