What does a fixed fee AI audit cost for a dental practice?

Four weeks, one fixed fee, one fixed scope, and nothing runs over. Week one is interviews. Week two we go through the software the practice already pays for. Week three we put a dollar figure and a build time against every fix we found and rank them. Week four you get the written plan and a walkthrough with whoever you want in the room. Payment is half at kickoff and half when the plan is delivered, so the second half is contingent on you holding the document.

What is not in the fee: no build, no software, no licence, no subscription, and no obligation to hire us afterwards. If you buy the audit and then build the top item with your own IT person or with a competitor of ours, you have used it correctly. The test of any audit is whether it still has value in a room we are not standing in, and dental practices get sold plenty of documents that fail it.

One practical note about the calendar. Four weeks is elapsed time, not your time. Interviews are short, they run one at a time, and we schedule them around chair time, which in most offices means before the first patient, during the lunch block, or after the last hygiene column closes. You get the full schedule before week one starts, so nothing lands on a day you needed the chair.

The owner dentist is the last interview, not the first. An audit built only from a management interview describes the office as it is understood upstairs, which is rarely where the work leaks. The people who know are the ones doing it, so the interview list in a typical practice is the front desk or scheduler, the treatment coordinator, whoever chases insurance verification and claims, the hygiene coordinator or whoever owns the recall list, the office manager, and one clinical voice, usually a hygienist or an associate. One at a time, and not in front of their boss.

We ask each of them the same unglamorous questions. What do you retype from one screen into another. What do you chase more than twice. What piles up on Monday. What did you do this week that you have done a hundred times before. What happens when you are at lunch, or out sick, or the phone rings while you are gloved up with a patient. Owners are often surprised by what comes back, which is the argument for asking rather than assuming.

Then we count. If the front desk says the phone gets away from them in the afternoons, that is a hypothesis, not a number, so we go and get the number: call logs, voicemail counts, the online booking form's submissions, the recall report, the open claims aging. Counting is what separates an audit from a workshop. It is also what makes the ranking defensible three months later when somebody asks why item two was not item one.

In a dental practice the list is fairly consistent: the practice management system, the phone system and its voicemail, the online booking widget on the website, whatever patient communication or texting tool is in use, insurance verification and the claims path, the website forms, the marketing spend and any call tracking attached to it, and the internal admin nobody calls a system, which is usually a shared inbox and a spreadsheet. We look at how they are actually used, which is often quite different from how they were sold.

Most practices run Dentrix, Eaglesoft, or Open Dental, and those systems differ a great deal in what they will let another tool read or write. Here is the honest boundary, and it matters: we do not publish a connector for any of them, and nothing in this audit is us telling you we already integrate with your system. What the audit does is establish which of three realistic approaches applies to your exact software and version, a live connection, a scheduled export the automation picks up, or a person exporting a file on a cadence, and what has to be verified before anyone promises anything. That verification is a demonstration against a test database with your version, not a slide. A vendor who answers yes to full integration before asking which version you run is describing a hope.

The audit needs to see your tools, not control them. Read access or a screen share with the person who uses the system is normally enough. On patient data, we work from counts and de-identified samples wherever the question can be answered that way, because the question is almost always how many and how long, not who. If any part of the work would put us near protected health information, we sign a business associate agreement first. To be explicit: no audit makes you HIPAA compliant, ours included, and nobody should tell you otherwise. Compliance describes how a whole setup is run, and it is a separate exercise with your own counsel.

Three questions worth putting to anyone who audits a dental practice, us included. Does the review look at your practice management system as it is actually configured in your office, or at a generic version of the product. Will any patient data leave the building during the work, and if so under what agreement and stored where. And does the recommendation depend on a product the auditor happens to sell. That last one is the tell. If the plan only works with the auditor's own subscription attached to the end of it, the plan is a sales route.

One row per job your team still does by hand. Each row carries what the job is, who does it today, how often it happens with the count we took, what fixing it is worth in dollars a year with the basis for that number stated, roughly what it takes to build, which instrument fits it, and what it depends on. The rows are ranked top to bottom, because an unordered pile of ideas leaves you exactly where you started, guessing which one to do first.

For a practice the rows tend to cluster in four places. The phone and everything that reaches it after hours. The recall and reactivation list, and whether anyone genuinely works it every week. Insurance verification and the claims that come back. And the quiet administrative sediment: intake forms retyped into the practice management system, referral letters, post-op follow up, the report the owner builds by hand every month. The instrument set against each row comes from the five things we build, which are workflow automation, voice AI, chat AI, AI agents, and AI visibility, and voice AI is one of them rather than the answer to everything. Some rows are not AI at all: a setting nobody turned on, or a policy that needs changing. Those go in the document too, because you paid for the truth rather than for a shopping list.

The roadmap also states what we would not do, and why. Sometimes the honest top row is fix the data first, or turn on the recall module you are already licensed for and see what it does for a month. If four weeks of counting concludes that nothing here pays back yet, the plan says so in its first paragraph and we say it to your face. That is a real outcome, and it is cheaper than the alternative, which is a build that solves a problem you did not have.

Ask any provider for one page of a real roadmap with the names redacted, ours included. You will learn more from that page than from an hour of slides. What you are looking for is whether the rows carry counts and dollar figures with a stated basis, or whether they carry adjectives. If every number is round and unlabeled, it was estimated in a proposal rather than measured in an office.

For a concrete sense of what one built row looks like once it is running, the dental work we can show publicly is a phone build. At My Smile Miami, month one produced 93 booked patients and roughly $27K in booked appointment revenue, both measured, and the office manager's own written account, client-reported rather than counted by us, is that Monday voicemail triage went from 45 minutes of waiting on the voicemail system to a summary with transcripts first thing in the morning. That was a voice build, not an audit finding. It is here so you can see the shape of a row that turned out to be worth doing, not as a forecast for yours.

If you already know your bottleneck, skip the audit. Plenty of owners can name it in one sentence: we miss calls at lunch and after six, or nobody works the recall list. You do not need four weeks of diagnosis to be told what you already said. You need a quote for that one fix, and we will give you one without selling you the audit first. Paying $4,500 to confirm your own diagnosis is the most common way this money gets wasted.

If the schedule is empty, the problem is demand, not operations. Making a quiet practice more efficient does not fill chairs, and that money belongs with whoever can fill them. Similarly, if your no-shows cluster around one provider, one time of day, or appointments booked six months out with no confirmation cadence, that is a policy question, and no amount of messaging will out-send it.

Wait if the practice is mid-change. A practice management migration in flight, an acquisition or DSO transition being negotiated, a new location opening in eight weeks, or a front desk that has turned over twice this year. A roadmap ranks stable work. If the work will not look the same in ninety days, the document ages before you can act on it, and you should call us after the dust settles rather than during.

If cash is tight, start free. Sit down with whoever answers your phone and whoever chases insurance, and write down every single thing they retype, re-enter, and chase for one week. That list is a rough first draft of what the audit formalizes, and it costs you an afternoon. We also run a free intake called the Free Opportunity Map: five fields, about two minutes, and within two business days an engineer writes back a ranked map of the three places AI or automation would pay back fastest in your practice, yours to keep. And there is a free thirty minute call. If the audit is not worth it for your practice, we will say so on that call, which is a cheaper conversation for both of us than the alternative.