A waste audit is a structured review of where a business loses money to repetitive manual work, rework, and revenue that never gets collected. A good one produces a ranked list of specific leaks with hours and dollars attached, not a general recommendation to adopt AI.
Last updated: July 2026
The term gets used loosely enough to be meaningless, so this page defines it, gives the checklist a credible one follows, and states the finding that vendors almost never put in writing: sometimes the correct answer is that you should fix your process and automate nothing.
What Is a Waste Audit Looking For?
Four categories, in rough order of how often they turn out to be the big one.
- Repetitive manual work. A human doing the same lookup, the same form fill, the same copy from one system into another, dozens or hundreds of times a week. Measured in hours per week and multiplied by loaded labor cost.
- Revenue that is never collected. Work that is not getting done at all because nobody has time. Unworked denials, dormant recall lists, aging AR past the point anybody is calling, quotes never followed up. This is usually larger than the labor category and almost always more surprising.
- Rework. Errors created upstream that get corrected downstream, repeatedly. Bad intake data producing denials. Wrong information entered once and fixed five times.
- Waiting. Work sitting in a queue because the next step requires a person who is busy. Claims waiting on a status check. Authorizations waiting on a submission. Files waiting on a signature.
Notice that only the first category is a labor cost. The other three are money, and they are usually the ones that justify doing anything at all.
What Does the Checklist Cover?
A real audit walks the operation function by function and asks the same questions of each. Here is the shape.
| Area | What gets counted | Typical finding |
|---|---|---|
| Intake and onboarding | Time per new patient or client, error rate, rework caused downstream | Manual data entry into two or three systems |
| Verification and eligibility | Lookups per week, minutes each, exception rate | Whole schedule verified by hand every morning |
| Scheduling and reminders | No show rate, confirmation labor, reschedule handling | Confirmations still done by phone |
| Claims and billing | Claim volume, denial rate by payer, unworked denial count, CARC distribution on the 835 | Denials discovered too late to appeal |
| AR and collections | Aging buckets, accounts nobody has touched, average days | A large bucket over 90 days with no owner |
| Recall and reactivation | Size of the dormant list, last contact date | A list nobody has worked in a year |
| Reporting | Hours spent assembling reports manually | Someone rebuilding the same spreadsheet weekly |
| Handoffs between systems | Number of places the same data is typed twice | The real source of the rework |
For each line the audit should capture four things: how often it happens, how long it takes, what it costs when it goes wrong, and whether the steps are actually the same every time. That last one determines whether it is automatable at all.
What Should a Waste Audit Produce?
If you finish an audit and what you have is a proposal, you did not get an audit. You got a sales call with a checklist stapled to it.
The deliverable should include:
- A ranked list of specific leaks, named concretely, with the volume and time attached to each
- A dollar estimate per item, with the assumptions stated so you can argue with them
- A clear split between what is automatable now, what needs a process fix first, and what should stay human
- A baseline measurement you can compare against later, agreed before any work starts
- An explicit note on which items are too small to be worth acting on
You should be able to take that document to a different vendor, or to nobody at all, and act on it yourself. If the findings only make sense if you buy something, they are not findings.
When Is the Right Answer “Automate Nothing”?
More often than the industry admits. Here are the honest outcomes.
Your process is the problem, not your capacity. If your denial rate is high because the front desk captures the wrong subscriber ID, automating denial appeals means processing self inflicted rework faster and more expensively. Fix intake. The denials go away rather than getting handled efficiently.
The volume is too low. A task that runs four times a week for ten minutes costs you about thirty five hours a year. No build pays that back. Automating it adds a system to maintain in exchange for almost nothing.
You already own the capability. A meaningful share of audits turn up features sitting unused inside the practice management system, CRM, or accounting software the business already pays for. Whether that is Open Dental’s recall module, an athenaCollector eligibility feed, Clio’s matter and billing reminders, or a QuickBooks AR aging report nobody configured, turning on a setting beats a build every time. Any auditor who does not check this first is not looking hard.
The data is not ready. If the underlying records are inconsistent, automating against them produces confident wrong answers at scale. Clean first.
Nobody can own it. Automation needs a human who reviews exceptions and notices when something silently stops working. If no one has fifteen minutes a week for that, the build will degrade quietly and you will not find out for months.
Any of those five is a legitimate audit result. A vendor who cannot deliver one of them has an incentive problem worth noticing.
How Long Should It Take?
An initial audit is a conversation, not an engagement. Thirty to sixty minutes with the person who actually knows how the work flows, plus a look at whatever volume data exists, is enough to identify the top three leaks and whether they are worth pursuing. A deeper audit, with real time and motion measurement, takes longer and should only happen after the first pass shows something worth measuring.
Be careful with anyone who wants to bill significant hours to tell you where your waste is. The discovery is the cheap part.
Where This Does Not Work
An audit is only as good as the data you can hand it. If you cannot produce claim volumes, denial counts by payer, AR aging, or a rough sense of hours per task, the findings will be directional at best. Two weeks of basic tracking before the audit makes the output far more useful.
It also does not work if the person in the room does not do the work. Audits conducted only with an owner or an office manager consistently miss steps, because the people performing the task have workarounds nobody documented. Get someone who touches the process daily into the conversation.
And an audit cannot resolve a disagreement about strategy. If the leadership team has not decided whether the goal is to cut cost, increase capacity, or improve patient experience, the ranked list will not settle it, because different goals rank the same findings differently.
How XClear Runs It
For context rather than as a pitch, here is one concrete implementation to compare others against. Our Waste Audit is free, takes 30 minutes, and produces the ranked list described above. If a build follows, we take 15 percent of documented savings. No savings, no fee. The automation works with the systems you already run, whether that is Dentrix, Open Dental, athenaCollector, eClinicalWorks, Clio, QuickBooks, or Salesforce, and all work touching PHI is handled under a Business Associate Agreement.
The revenue cycle version of this is described on our medical billing page, and the broader approach is under solutions.
FAQ
What is a waste audit? A structured review of where a business loses money to repetitive manual work, rework, waiting, and revenue that never gets collected. It should produce a ranked list of specific leaks with hours and dollars attached.
How long does a waste audit take? An initial audit is typically 30 to 60 minutes with someone who knows how the work actually flows. Deeper time and motion measurement takes longer and should only follow a first pass that found something worth measuring.
What should a waste audit cost? An initial audit should be free or inexpensive. Discovery is the cheap part of this work. Be cautious about paying significant consulting hours simply to be told where your waste is.
Can a waste audit conclude that we should not automate anything? Yes, and a credible one sometimes does. Common reasons include a broken upstream process, volume too low to pay back a build, unused features in software you already own, and data that is not clean enough to automate against.
What do I need to prepare for a waste audit? Rough volumes for your highest frequency tasks, denial or error counts if you have them, AR aging if relevant, and access to someone who performs the work daily rather than only someone who manages it.
If you want the ranked list for your own operation, book the free 30 minute Waste Audit. If the finding is that you should fix a process instead of buying automation, that is what the document will say.