DSO Day 1–100 integration: same-store, add-backs, PMS migration.

01 Same-store baseline — locked in week one, not month three

Same-store performance is the financial scoreboard for the next four quarters of the integration and the next four years of the platform. The buyer needs the as-acquired baseline locked in week one, before any post-close change can muddy the read. Every conversation about post-close performance — earn-out math, lender covenants, LP letter — anchors to this baseline. If the baseline is not defensible, every conversation about performance becomes a conversation about accounting.

The canonical definition we use across DSO, vet, and derm integrations is the one the franchise and multi-unit retail world has used for decades: locations continuously operating for at least 12 full months before the base period, in a stable capacity configuration, with comparable provider availability and no major service-line changes that fundamentally alter case mix. That definition runs three layers — seasoning, capacity, service mix — and each layer has to be locked explicitly. Some sponsors use 18–24 months of seasoning for slow-ramp specialties (ASCs, higher-ticket implant practices). The number is less important than the documentation. Document what you used, attach it to the integration charter, and stop arguing about it.

The seller-to-buyer convention bridge

Every DSO acquisition arrives with a same-store reporting convention that the seller built for their own management team — and almost never matches the buyer's portfolio standard. The seller counted sites after six months. The buyer counts after twelve. The seller treats a chair addition as same-store; the buyer pulls it out and amortises. The seller reports revenue gross of refunds; the buyer reports net. In week one of every integration we run, the finance lead rebuilds the trailing 24–36 months under the seller's convention, then a parallel version under the buyer's convention, and reconciles the two with a named bridge. The bridge has three columns — seller convention, buyer convention, named adjustment — and a one-line owner per row. That document is the financial spine of the next four quarters.

The five pitfalls that move the read

Five accounting choices drive most of the noise in same-store reporting, and each of them has to be explicit in the bridge. Provider mix changes — calling growth same-store when it is driven by adding a high-productivity associate or losing a low-productivity one. The fix is to track and disclose revenue per provider day and visits per provider day, decomposed into volume per provider day, provider days growth, and revenue per visit. Fee schedule resets — a 10–15% fee increase reported as same-store growth without the volume decomposition behind it. The fix is to maintain an index of standard fee schedules and report a constant-price same-store view. Payer contract renegotiations — same-store jumps from a payer mix shift, not from demand. The fix is to report payer mix by revenue and visits, and allowed amount per unit by payer category. Calendar quirks — comparing quarters with different numbers of clinic days. The fix is to standardise on per-clinic-day metrics. And capacity expansions — growth from added operatories reported as same-store. The fix is to maintain the location in the cohort but provide a capacity-expansion bridge showing growth from utilisation versus growth from capacity. All five live in the bridge. None of them live in the headline number.

02 Doctor compensation and the add-back recovery curve

Doctor compensation is the variable that most often drives an LOI re-trade and the variable that most often blows up in the first 100 days. Two conversations are live at the same time post-close, and the integration playbook has to keep them on different clocks. The first is the deal-time add-back conversation — the owner doctor ran personal expenses through the practice, paid family on the books, took CE that read like discretionary spend, and the LOI EBITDA was struck on the assumption that the buyer would remove all of it. The second is the harmonised compensation model — the new ownership has a standard structure (cash + rollover equity, productivity-weighted comp, quality modifiers) and every selling doctor has to be re-recruited under it. Both conversations need to happen. Neither one should happen on Day 1.

What add-back recovery actually looks like

The published advisor commentary and the practitioner conversations all converge on a similar curve. First 30 days post-close: roughly 10–30% of LOI-stage add-backs realised. The new owner does not want to shock staff or patients in the first month, and the seller is still mentally in charge. First 60 days: 30–50% cumulative recovery as the new employment agreement takes hold, some vendor contracts get rationalised, and centralised services start to land. First 90 days: 40–60% cumulative recovery is typical, with 70%+ achievable in well-managed integrations where the selling doctor has rolled meaningful equity and is leaning into the work. Conservative, well-documented add-backs can climb to 80–90% over twelve months. Over-aggressive or weakly documented add-backs deliver 0–20% true recovery and become the conversation that sours the post-close relationship.

Why add-backs come in at 50%, not 100%

The most common buyer mistake is treating the LOI add-back schedule as a recoverable cash flow and the most common seller mistake is treating it as defended EBITDA. Neither is true. Add-backs come in at 50%, not 100%, for four reasons we see in nearly every integration. First, some "perks" are embedded in operations. The owner takes more CE than an associate would because the owner does the treatment planning. The owner pays the long-tenured front-desk lead above market because she holds the patient relationships. Cut comp and production falls. Second, owner comp normalisation breaks down behaviorally. The LOI assumes the owner cost gets cut to a 30%-of-collections associate rate. Post-close, if the seller keeps producing like an owner, the DSO ends up paying above the modeled rate via bonuses and benefits. If the seller slows down, the DSO has to backfill with associates at 30%+ collections — wiping out the savings. Third, transition friction is real. Cutting staff or hours in the first 60 days causes patient attrition, lower case acceptance, and front-desk turnover. Fourth, integration cost. IT, central billing, training, payroll consolidation — each eats into near-term savings before the run-rate stabilises.

The first-90-days run-as-is thesis

Across the 7 DSO integrations in our sample, the pattern that distinguished the platforms that hit underwriting from the ones that missed it was the discipline of running the first 90 days as-is on compensation. Doctors run on their pre-close comp through Day 90. The harmonisation conversation begins in month four, with the same-store baseline locked, the integration team's competence demonstrated, and the clinical leadership's trust earned. The PE sponsors who model "operational EBITDA" at 6–12 months rather than "paper EBITDA" at LOI are the ones who exit at premium multiples; the ones who model full add-back recovery in month three are the ones who write to LPs about "stabilisation periods" and "integration headwinds" twelve months in. The math is on the side of patience.

03 PMS unification — the loudest mistake to defer

Practice management system unification is the loudest harmonisation project in any DSO integration and almost always the wrong one to lead with. The cost of the migration is not the implementation fee. The cost is three months of half-trained front-desk and clinical teams running production under a new system — and the production they do not collect during that quarter is non-recoverable if it lands inside the trust-building window with the selling doctors and the front office. Eaglesoft to Dentrix Ascend, Open Dental to Curve, Dentrix to Carestack, anything to Denticon — each migration carries a conversion-quarter cost. The sequencing decision is the work.

The conversion-quarter drift band

A defensible planning band for production drift in the PMS conversion quarter is 8–18%, with the median in our DSO sample landing near 12%. The drift is not chair time. It is the compounded loss of slower check-in and checkout, scheduling friction, staff unfamiliarity with the new templates, incomplete imaging workflows, claims and eligibility hiccups, and delayed posting and follow-up. A/R aging often worsens 10–30 days in some buckets during the first quarter because attachments and follow-up cadence slow down. Claim rejection rates spike in the first 1–2 billing cycles as payer mappings get tested in production. Front-desk performance recovery typically takes 2–6 weeks if training started 2–4 weeks before go-live, and longer if it did not. Full stabilisation is rarely complete inside the conversion quarter.

Platform-by-platform — what we actually see

Across our 7 DSO platforms, the migration complexity has been consistent enough to generalise. Eaglesoft migrations are the most operationally sensitive — long-tenured desktop footprint, imaging workflows that almost always need extra care, downstream data export that costs more time than the budget assumed. Dentrix Ascend is moderate complexity — cloud-based and designed for multi-location, but reporting and customisation have a real learning curve and "guided conversion" is not the same as frictionless. Open Dental is moderate but punishing if discipline slips — strong API-based ecosystem, but unauthorised direct database writes from third-party tools have caused multi-week recovery episodes in two of our platforms. Curve Dental is moderate — cloud architecture reduces infrastructure pain but workflow mapping still costs. Carestack is moderate-to-high — enterprise-oriented, with more modules and more potential integration points, risky if too many modules turn on at once. Denticon is moderate — multi-location-native, but multi-site standardisation is the hard part, not the software.

The sequencing rule we use

PMS migration in month four or five at the earliest, never in the first 90 days. The order across the systems runway is: core infrastructure readiness (SSO, network, imaging interfaces, backup/restore, user provisioning) first; then PMS migration with the schedule, patient data, financials, insurance, codes, basic reporting; then revenue-cycle stabilisation (claims, eligibility, payment posting, AR cleanup, denial management); then adjacent systems last — CRM, AI receptionist, marketing automation, BI dashboards. The rule of thumb that runs the whole sequence: do not stack major integrations during the cutover window. If something breaks, the team has to be able to tell whether the root cause is the PMS, the integration layer, or the downstream tool. Compound failure is the integration killer because it is unattributable, and unattributable failure spreads blame.

04 The systems runway after PMS

Once PMS is done, the systems runway opens for the rest of the integration. Billing centralisation, supply-chain consolidation, lab vendor rationalisation, brand alignment, chart-of-accounts harmonisation, payroll consolidation, benefits standardisation. Each item has a window; the integration team's job is to sequence them so no single location is taking more than two changes inside any 90-day window. The same logic that protects the clinical team from PMS in week three protects them from compound system change in any quarter of year one.

The order we run across operating-platform integrations is roughly: chart of accounts and management reporting in month two (the integration team needs to read the books), payroll and benefits consolidation in month three (cost-of-living and benefits anxieties have to land cleanly), PMS migration in month four or five, central RCM bolt-on starting in month six, supply-chain and lab consolidation in months seven through nine, brand alignment in month ten or later if at all. Brand alignment is the most overrated early initiative — patients almost never care, and the cost of changing signage and patient communication during the trust-building window is high relative to the value. Defer it unless it is in service of a near-term capital event.

The two-change-per-quarter rule

The discipline that has held across every successful integration in our sample is the two-change-per-quarter rule: no single location takes more than two material operational changes inside any 90-day window, measured at the location level not the platform level. The platform integration plan can have eight changes in flight at once across 22 locations as long as no single office is exposed to more than two. The corollary is that the integration calendar is built bottom-up by location, not top-down by workstream. The PE sponsor wants to know when the platform-level milestones land; the integration leader wants to know which locations are over the two-change cap in which weeks. Both views matter; the location view is the one that prevents the doctor and front-desk attrition that kills the LOI.

05 Integration management office cadence and decision rights

Every successful integration runs an Integration Management Office (IMO) that exists from Day −30 through Day 200. The IMO has five named owners — integration lead (usually a senior ops executive on the buyer side), finance lead, IT/PMS lead, HR/people lead, clinical lead. The IMO runs a single shared tracker, a decision log, and a risk register from Day 1. The cadence is daily huddles in week one, then weekly through Day 60, then biweekly through Day 100. Skip the IMO and the integration runs on hallway conversations; the conversations that needed to be hard get made by the path of least resistance, and the path of least resistance is almost never the path of best EBITDA.

The single most useful artifact the IMO produces is the decision log — a one-row-per-decision document that captures the date, the decision, the owner, the rationale, the open risks at the time. In every integration where the LP letter needed to explain a same-store miss or an add-back shortfall, the decision log was what allowed the platform CFO to defend the trajectory. In every integration where the decision log was not kept, the conversation became "we did everything we could" — which is not a financial answer and not a defensible one.

What the IMO owns and what it does not

The IMO owns the integration calendar, the change-management protocol (the two-change-per-quarter rule), the decision log, the risk register, the integration scorecard, and the escalation path to the platform CEO and PE sponsor. The IMO does not own clinical decisions, compensation negotiations with individual doctors, or vendor selection for clinical equipment — those stay with the clinical leadership and the platform leadership. The dividing line matters because every IMO that drifts into clinical decision rights loses the trust of the clinical team in the second quarter; every IMO that holds the operational line keeps it.

06 Three questions to ask the buyer at close

If a selling doctor is reading this in the run-up to close, or a sponsor is evaluating an integration team before signing the LOI, these are the three questions that separate the platforms that hit underwriting from the ones that explain why they did not.

1. 01
   Is the same-store baseline locked under both the seller's and the buyer's conventions, with a named bridge between them, by Day 14? If the answer is "we are working on it" or "we will get to it after the IT cutover," the integration is starting on financial sand. Every same-store conversation for the next 36 months will be a conversation about accounting drift instead of operating performance. Lock the bridge in week one.
2. 02
   When does the doctor compensation harmonisation conversation begin — week three, month four, or "we have not decided"? The right answer is month four. Week three is too aggressive and breaks the run-as-is thesis. "We have not decided" is worse than wrong — it tells the selling doctor that the integration plan does not exist in writing, which is the strongest predictor of doctor attrition in year one. The compensation conversation is the most consequential conversation of the entire integration; it should have a date.
3. 03
   Where does PMS migration sit on the systems runway, and has anyone modelled the production drift in the conversion quarter? If the answer is "we are doing it in month two," walk. If the answer is "month four or five, we have modelled 8–18% production drift in the conversion quarter and structured the financial plan around it," the integration team has done the work. The willingness to name the drift band is the diagnostic.

07 What we watch through Day 200

The integration is not done at Day 100. Day 100 is the end of the trust-building window and the beginning of the platform-build window. We watch four signals through Day 200 to decide whether the integration is on track. First, same-store revenue and visit growth versus the locked baseline, by location, not by platform — platform-average masks the location-level dispersion that matters. Second, doctor retention measured at the named-doctor level, not the FTE level — losing a 4-day-a-week associate and replacing with a 4-day-a-week associate is not net-zero on the same-store read. Third, A/R aging trend and claim rejection rate, particularly post-PMS migration — these are the leading indicators of cash conversion that lender covenants test. Fourth, integration cost actuals versus budget, by workstream — most integration plans run 30–50% over budget in year one, and the platforms that catch the overrun in month four are the ones that finish year one within the original underwriting case.

For the buyer, the work that pays for itself most reliably in this window is rebuilding the location-level operating P&L and circulating it to every location lead by month two. The median first-100-day same-store contribution lift in our DSO sample, across 7 platforms, was approximately $2.1M on a 6–28 location base — almost entirely from rebuilding the location-level visibility that the seller's management reporting did not provide. The contribution comes from finding the underperforming half-day of hygiene in the schedule, the high-margin procedure mix that the seller stopped marketing, and the working-capital lag that the seller had absorbed but the buyer can collect. The pattern repeats; the magnitude is what varies.

08 The sequence, end to end

The DSO integration sequence we run across operating platforms is short enough to fit on a page and disciplined enough to defend in front of the LP. Day −30 through Day 1: stand up the IMO, name the five owners, build the Day 1 script for staff and patients, lock the integration charter. Week 1: business continuity first, nothing changes that touches the patient or the paycheck. Days 1–14: rebuild the same-store baseline under both conventions, build the seller-to-buyer bridge, lock it in writing. Days 14–60: chart of accounts harmonisation, payroll and benefits consolidation, location-level operating P&L circulated to every location lead. Days 60–100: PMS decision and pilot environment built, doctor compensation harmonisation framework drafted with the clinical leadership, 1:1 conversations scheduled. Days 100–150: PMS migration window opens for the first cohort of locations, doctor compensation conversations begin with documented before/after math. Days 150–200: PMS rollout to remaining locations, central RCM bolt-on, integration scorecard reported to the sponsor monthly. Each gate has an owner; each gate has a date; the calendar is built bottom-up by location. The integration team's product is the sequence — not the items.