DSO IT Spend Runs 2.4% of Collections. 1.5% Is Enough

On a practice collecting $1 million a year, the difference between an IT budget of 2.4% of collections and one of 1.5% is $9,000 a year, every year, forever, and almost none of it shows up in production. That $9,000 is the quietest line in your overhead because nobody renegotiates it: software renews on a card, seats get added when you hire, and the invoice arrives in twelve pieces small enough to ignore.

Here is the claim an informed owner can fight us on: most of what a group practice spends above 1.5% of collections on technology buys coordination across locations, not clinical quality or collections, and a single-site owner who copies the DSO stack is paying to solve a problem they do not have. A DSO buys software to make forty practices behave like one. You own one.

The Ratio Nobody Actually Publishes

Start with what is real. There is no audited, industry-wide figure for DSO IT spend as a share of collections. What exists is a range for technology and equipment inside dental overhead, and the published versions of it disagree by a factor of three.

Financial Advisors for Dentists, citing the ADA Health Policy Institute 2023 Survey of Dental Practice, puts technology expenses at 1% to 3% of collections. Overjet, drawing on ADA HPI plus PKF O'Connor Davies benchmarks, puts equipment and technology at a healthy 3% to 5% and red-flags anything above 7%. Both are defensible, because they are counting different things. The narrow number is software, support, security and connectivity. The wide number folds in equipment ownership: chairs, handpieces, sensors, mills, the capital side.

That distinction is the whole argument. We are talking about the narrow number: the recurring, contractual, per-seat, per-location, auto-renewing spend. Call the group-practice version of it 2.4% of collections. That is our working figure for the argument, built from the top of the software range plus the integration layer groups add on top, not a survey result anyone has published. The point of the number is not its decimal place. The point is that the gap between a disciplined single-site budget and a group-practice budget is roughly a percentage point of collections, and a percentage point of collections is real money in a business where the whole overhead structure gets fought over in half-points.

What $9,000 Buys and What It Doesn't

Use the middle of the market we can actually measure. Across the Practice Market Index, The Dental Signal's dataset of dental practices currently listed for sale, median collections are $1,000,000 (n=850) and the median practice has 5 operatories (n=851). Collections per operatory run a median of $200,639 (n=728). These are practices on the market, the asking side, not a cross-section of all practices, but they are a fair picture of the mid-market operation most owners recognize.

Run the two budgets against that practice.

• At 2.4% of collections, IT costs $24,000 a year, or $4,800 per operatory.
• At 1.5%, IT costs $15,000 a year, or $3,000 per operatory.
• The gap is $9,000 a year, $45,000 over five years, before any contractual price increases.

Now put it where owners feel it. A practice running 17 patient visits a day across roughly 245 clinical days sees about 4,165 visits a year. At 2.4%, technology costs $5.76 per visit. At 1.5%, it costs $3.60 per visit. Two dollars and sixteen cents a patient. That sounds like nothing until you ask what else you could buy for $2.16 a patient: a second hygiene recall touch, a real post-op call, a printed treatment plan the patient actually takes home.

Here is a version of the same budget built to 1.5% on a $1 million practice, and it is not austere:

• Practice management software, claims and e-statements: $7,200
• Imaging software licenses and sensor service plan: $2,400
• Backup, endpoint security, patching and a real monitored firewall: $2,100
• Hardware replacement reserve, funded monthly: $2,400
• Website, hosting, domain, forms: $900

That totals $15,000. Everything clinically necessary is covered, security is funded rather than hoped for, and there is a sinking fund so the next five-year computer refresh is not an emergency on a credit line. What is missing from the list is the layer that pushes a budget to 2.4%: multi-location reporting dashboards, a second communications platform that duplicates what the practice management system already sends, an AI add-on bought on a demo, a phone system with per-seat analytics nobody opens, and the consultant retainer that exists to make those four pieces talk to each other.

What Most Owners Get Wrong: Buying the Integration Layer

The specific mistake is this: owners of one or two locations buy the software a 40-location group needs, because that is the software being marketed at the meetings they attend and the software their acquisition-minded peers brag about.

Group practices have a genuine problem. They need one source of truth across sites that use different staff, different fee schedules and sometimes different practice management software. They are also buying with a resale narrative in mind: standardized reporting makes a platform easier to sell. That justifies spending real money on the connective tissue between systems. It is not vanity for them. It is the product.

A single-site owner with 5 operatories has none of that. The source of truth is the schedule on the wall and the person who runs it. When that owner buys the group-practice stack, three things happen. The spend goes up by about a point of collections. The front desk now works in two systems that each hold half the patient's history. And the one metric that actually drove the purchase, usually unscheduled treatment or hygiene reappointment, moves by less than the cost of the software, because the constraint was never data visibility. It was that nobody owns the follow-up call.

Our view: if you cannot name the person whose weekly behavior will change because of a new platform, the platform is overhead with a login screen. We would not sign it.

The counter-mistake is just as expensive and it is worth being fair about it. Underspending is a documented condition in this field, not a theoretical one. A peer-reviewed study of electronic dental record adoption found 52% of surveyed U.S. practices had adopted an EDR by 2012, with cost cited as the top barrier by 28% of non-adopters. Plenty of practices are not running lean; they are running on a 2014 server in a closet with no tested backup. That is not 1.5% discipline. That is deferred maintenance with a breach waiting at the end of it.

Per-Seat Pricing Is Erosion, Not Drift

Technology spend deserves the Drift vs. Erosion test more than almost any other overhead line, and most owners misclassify it.

Drift is a vendor raising a $300 monthly fee to $320. Annoying, recoverable, manageable operationally. Erosion is a contract whose price is indexed to something that grows: providers, operatories, locations, patient records, text messages sent. That structure means your software bill rises every time the practice succeeds, and it rises whether or not collections rise with it.

Run the compounding. A $15,000 budget with a 7% annual escalator costs $19,661 in year five. If collections stay flat, the ratio has walked from 1.50% to 1.97% without a single new purchase decision. Add one associate at two per-seat platforms and you cross 2.4% having made no choice you would remember.

That is erosion, and it is not fixed by shopping harder. It is fixed by renegotiating the pricing structure: annual caps written into the renewal, flat-practice pricing instead of per-provider, or a consolidation that removes a vendor entirely. Treating erosion as drift, absorbing it quarter by quarter because each increase is small, is the expensive mistake. The escalators compound. Your fee schedule does not.

When We'd Spend Above 1.5% Without Blinking

We are not arguing for a ceiling on ambition. We are arguing for a default. The ADSO white paper cites an ADA survey finding 49% of dentists plan to spend more on technology over the next two years, and that direction is right as long as each increase has to earn its place.

The test is simple and it has a number in it. A tool at $1,000 a month costs $12,000 a year. One additional crown a month at a $1,300 fee produces $15,600 a year. If a diagnostic or treatment-presentation tool reliably adds one case a month, it pays for itself and clears about $3,600. That is a real yes. We would sign that, and we would also put a 90-day review on the calendar to confirm the extra case actually appeared in the production report rather than in the vendor's case study.

What we would not do is let capital equipment hide inside the IT line. A scanner, a mill, a 3D printer and new sensors belong in the equipment budget, where they get evaluated on payback period and financed deliberately. Blending them into "technology" is how a practice ends up reporting 5% and genuinely not knowing whether that represents productive assets or a stack of subscriptions.

The Audit That Takes One Evening

Pull twelve months of card and bank statements and highlight every recurring technology charge. Owners who do this for the first time typically find three things: a subscription for a product nobody on staff can demonstrate, a per-seat charge for a provider who left, and two systems sending the same patient the same reminder.

Then compute one number: total recurring technology spend divided by collections. If it is under 1.5% and your backups are tested, you are fine and you should stop reading the vendor emails. If it is between 1.5% and 3%, name the outcome each line item produces. If it is above 3% and you run one location, you are funding a group practice's problem out of a single practice's margin.

What Would Change Our Mind

Two things. First, credible per-practice data showing that AI diagnostic and revenue-cycle tools lift collections by more than their cost at single-site scale, measured in production reports rather than vendor decks. If a $12,000 platform reliably returns $40,000, the right budget is not 1.5% and we will say so plainly. Second, insurer behavior. If payers start requiring electronic attachments, real-time eligibility and structured clinical data as a condition of timely payment, the floor rises for everyone and the 1.5% default becomes a liability instead of a discipline.

We are watching the renewal pricing, specifically. The per-provider model is spreading, and a contract that taxes your growth is a worse deal than a flat contract that costs 20% more today. Anyone handing us a per-seat escalator with no cap is quoting a number that is not the number.

Do This Before Your Next Renewal

This week, list every technology vendor you pay, the monthly amount, the renewal date and the name of the staff member whose job depends on it. Cancel anything with a blank in that last column. For your two largest contracts, email the account manager and ask for flat-practice pricing with a capped annual increase before the auto-renew date, because that ask only works before the renewal, never after. Then set a calendar reminder for the same audit twelve months out, because this line does not stay fixed on its own.

If you want to see where your technology spend sits against the rest of your cost structure rather than in isolation, put your collections and your full expense breakdown into our free overhead benchmark and run your own numbers. It will tell you whether that one point of collections is your real problem or just the one that happens to arrive by email every month.