TL;DR: A practice handling 40 to 50 inbound faxes a day spends roughly 4 to 12 staff hours daily working that queue, and EHR integrated fax triage and routing typically recovers 60% to 80% of it once the model is tuned. The savings come from two places: documents that file themselves with no human touch, and exceptions that drop from a full manual workup to a short review. Your real number depends on daily volume, document mix, and the loaded hourly cost of whoever works the queue.
What your fax queue costs before anything is automated
Start by measuring what you have, not what a vendor tells you the industry average is. Every inbound fax carries the same chain of manual steps: someone opens it, figures out what kind of document it is, matches it to the right patient, indexes it into the chart under the right document type, routes it to the provider or work queue that needs to act, and sometimes creates a task so it doesn't die there.
Those steps take wildly different amounts of time depending on the document. A clean lab result for an established patient might be a two-minute touch. A 40-page hospital discharge packet for a patient your system has never seen, arriving as a crooked scan, can eat 15 minutes before anyone acts on the clinical content.
You'll see numbers like "12 to 15 minutes per fax" quoted across the internet. Most of those come from vendor marketing and none of them are your practice. Spend one week having two staff members log a timestamp when they pick up a document and when they finish it, tagged by document type. That single week of data is worth more than every benchmark you'll find, and it's the number a CFO will actually accept.
Fax volume isn't going away on its own, either. The most recent federal data on hospital interoperability found that only 43% of non-federal acute care hospitals routinely engaged in all four domains of electronic exchange in 2023 — and among independent hospitals, that figure was just 22%. If a meaningful share of your referral sources still can't route a record electronically, your queue is structural, not temporary.
The five inputs behind any honest fax automation ROI model
Every defensible model for EHR integrated fax triage and routing runs on five variables. If a savings calculator doesn't ask you for all five, it isn't modeling your practice — it's modeling a brochure.
- V — daily inbound volume, split by document type. Not one blended number. Labs, referrals, records requests, and payer correspondence behave differently at every step.
- T — current average minutes of staff touch per document, by type. From your own timing study.
- A — automation rate. The share of documents in each type that clear end to end without a human touch, at the confidence threshold you set.
- R — residual review minutes. What the exceptions still cost after automation. This is not zero, and modeling it as zero is the single most common way these business cases get inflated.
- C — loaded hourly cost of the staff working the queue.
The arithmetic is simple once you have the inputs:
- Hours per day today = Σ (V × T) ÷ 60, summed across document types
- Hours per day after = Σ (V × (1 − A) × R) ÷ 60, plus your QA sampling time
- Annual savings = (hours today − hours after) × operating days × C
Run it per document type, then sum. A blended average will overstate savings on your hard documents and understate them on your easy ones, and the error compounds.
This is the category Honey Health's Fax Triage agent sits in — software that classifies the inbound document, matches it to the patient, and files and routes it inside the EHR without a person in the loop. Whichever platform you evaluate, insist on running your own numbers through the formula above rather than accepting a vendor-supplied output. A model you built is a model you can defend to your board.
How do you calculate loaded hourly cost without fooling yourself?
Base wage is the starting point, not the answer. According to BLS Occupational Employment and Wage Statistics, medical secretaries and administrative assistants working in offices of physicians earned a mean hourly wage of $19.92 as of May 2023, with the national mean across all settings at $20.85.
Then add benefits. BLS Employer Costs for Employee Compensation put wages and salaries at 70.3% of total employer compensation cost for private industry workers in March 2025, with benefits making up the other 29.7%. Dividing a $19.92 wage by 0.703 gets you roughly $28.34 per hour in loaded cost.
Two more adjustments matter:
- Productive hours, not paid hours. PTO, training, and meetings mean an FTE delivers closer to 1,800 productive hours a year than 2,080. If you're converting saved hours into FTE equivalents, use the smaller denominator.
- Use the right person's wage. If your fax queue is worked by an RN triaging referrals or a biller chasing payer correspondence, the front-desk wage understates the cost badly. Model each document type at the cost of whoever actually touches it.
A worked example, clearly labeled as illustrative
The numbers below are illustrative inputs chosen to show how the model behaves. They are not measured results from any deployment, and you should replace every one of them with your own.
Take a 14-provider multi-specialty group receiving 120 faxes a day:
- 45 clinical results and labs at 4 minutes each
- 30 referrals with attachments at 12 minutes each
- 25 records requests and correspondence at 6 minutes each
- 20 payer documents (auth responses, remits) at 9 minutes each
Baseline: (180 + 360 + 150 + 180) ÷ 60 = 14.5 staff hours per day.
Now apply automation rates of 85% on results, 50% on referrals, 75% on records, and 60% on payer documents, with residual review of 2, 6, 3, and 5 minutes on what doesn't clear, plus a 5% QA audit of auto-cleared documents at one minute each.
After: roughly 162 minutes of exception review plus 4 minutes of QA — about 2.8 staff hours per day.
That's 11.7 hours a day recovered, or about 81% of baseline. Across 250 operating days at $28.34 loaded cost, that's roughly 2,900 hours and $83,000 a year — call it 1.4 FTE worth of capacity.
Note that 81% sits at the optimistic end of the range in the TL;DR. If you're building the business case yourself, haircut year one by 20% to 30% and let the upside be a pleasant surprise rather than a missed forecast.
What the raw hours number hides
The hours figure is the headline. It is also the number most likely to get you in trouble in month six if you present it without three caveats.
Hours saved are redeployed capacity, not automatic payroll reduction. You rarely eliminate 1.4 FTE. What you get is 1.4 FTE of capacity back, and the value depends entirely on what you do with it. If you're growing, it's a hire you didn't make. If you're flat, it's overtime you stopped paying and a backlog you finally cleared. If you genuinely intend to reduce headcount, put that in writing in the business case, because savings that nobody acts on show up as zero on the P&L.
Month one costs real time. Someone experienced on your team has to define routing rules, map document types to your chart structure, name the work queues, and correct mismatches while confidence thresholds settle. Budget 20 to 40 hours of that person's time in the first month. Vendor calculators almost never include this, and it's the line item that makes your model credible to a skeptical CFO.
Accuracy ramps over a quarter, not a day. The easy document types perform well early. Multi-page hospital packets, handwritten intake forms, poor-quality scans, and patients with no prior record in your system improve as the system sees more of your specific referral sources. Model the first 90 days at a lower automation rate and step it up.
What else is worth quantifying besides hours?
Hours are the easiest thing to count and rarely the most valuable thing you gain. Three second-order effects deserve their own line in the analysis.
Referral response time. Measure hours from fax arrival to first patient contact attempt, before and after. A referral that sits in a queue for three days is a referral that may get scheduled somewhere else. Multiply the change in your conversion rate by your average revenue per new-patient episode and you have a number that usually dwarfs the labor savings.
Documents that never get worked. Pull the age distribution of your current queue. If anything is older than 72 hours, you have leakage you're not currently pricing. Automated routing doesn't just move documents faster — it removes the failure mode where a document is never assigned to anyone at all.
Deadline exposure on payer correspondence. Prior auth responses and payer requests carry clocks. The AMA's 2024 survey found practices completing an average of 39 prior authorization requests per physician per week, with physicians and staff spending about 13 hours a week on them and 40% of physicians employing staff who work on prior auth exclusively. A payer document that sits unrouted for four days is a denial risk with a dollar value you can estimate from your own appeal rate.
Also worth noting: an agent works the queue overnight and on weekends. Monday morning starting at zero instead of at a Friday backlog is a scheduling benefit that never shows up in an hours-saved calculation.
Frequently Asked Questions
How many faxes a day does a practice need before automation pays for itself?
There's no universal floor, but the math turns on total touch minutes, not document count. A practice with 40 faxes a day that are mostly complex referrals can clear a higher annual savings figure than one with 120 simple lab results. Run your own volume and time inputs through the five-variable model before assuming you're too small.
Does EHR integrated fax triage and routing work if my EHR has no open API?
Usually, yes, though the integration path changes. Many platforms write into the chart through HL7 interfaces, document-management endpoints, or supervised interface automation when no modern API exists. Ask any vendor specifically how they'll write to your EHR, who maintains that connection, and what happens when your EHR ships an update. Vague answers here are a real risk signal.
What automation rate should I actually expect in year one?
It depends heavily on document mix. Structured, recurring documents from consistent senders tend to clear at high rates within weeks. Unstructured multi-page packets, handwritten forms, and documents for patients with no record in your system take longer. A reasonable first-year plan assumes strong performance on your top two or three document types and modest performance on the rest.
Will this let me reduce front-office headcount?
It can, but that should be an explicit decision rather than an assumption baked into the model. Most practices convert the recovered hours into faster referral turnaround, eliminated overtime, or avoided hiring as volume grows. Decide which of those three you're actually buying before you present the number, because each one shows up differently on the P&L.
How do I pressure-test a vendor's savings claim?
Ask three questions: what residual review time did you assume on exceptions, what automation rate did you assume by document type, and what implementation time did you charge to month one. A calculator that assumes zero residual review and instant accuracy will produce a number roughly double what you'll see. If a vendor won't show the inputs, treat the output as marketing.
