How the number breaks down
Methodology summary
Four views, two axes
Every practice scored on Revenue drivers (demand, capture, conversion) and EBITDA drivers (expense ratios, throughput, waste) from two independent sources (PBI financials, lagging; Dental Intel operations, leading).
Lever math
Revenue levers: captured revenue × marginal margin (avg+25pp, cap 65%) × 30% realization, capped at 8% of TTM.
Expense levers: expense-ratio gap × revenue base × 30% realization (flows 1:1 to EBITDA).
Empirical validation
24 months of DI history × 13 metrics tested for within-practice predictive power. Strongest leading indicator: grossHygieneProduction (|r|=0.62 @ lag 3). Methodology tab has full matrix.
Divergence watch — where leading signals are telling us something PBI can't yet see
A practice's DI score (leading) diverging from PBI score (lagging) is often the earliest warning of a margin move in the next 90 days. Here are the 10 largest divergences in the network.
| Practice | ROD | PBI Rev | DI Rev | PBI EBITDA | DI EBITDA | Signal |
|---|
Top 10 $ opportunities
| # | Practice · ROD | TTM Rev | Margin | Revenue Driver | Rev $ | Expense Driver | Exp $ | Combined $ |
|---|
By Regional Ops Director
| ROD | Practices | TTM Revenue | TTM EBITDA | Margin | Top-lever $ (12mo) |
|---|
Coverage disclosure
What's next
- Monthly refresh — move from snapshot to automated month-end pipeline, feeding both DI + Gen4 PBI
- Complete INTACCT P&L coverage — currently 84 practices; adding the remaining ~180 unlocks practice-level expense-ratio levers network-wide
- Correlation re-run — with 36+ months of DI history and richer features (YoY change, trend z-scores) strengthens the empirical signal
- ROD drill-down dashboards — ROD-specific rollup with their own top-10 leverage list, exportable
4-View Comparison Grid
Every practice scored four ways: Power BI Revenue, Power BI EBITDA, Dental Intel Revenue (leading), Dental Intel EBITDA (leading). Divergences surface practices where leading indicators drift from lagging financials — the insight surface.
| Practice · ROD | PBI Rev | PBI EBITDA | DI Rev | DI EBITDA | TTM Rev | TTM EBITDA | Margin | Divergence | Top Lever $ |
|---|
EBITDA Leverboard
Every practice ranked by top-lever 12-month EBITDA upside, with projected margin delta after execution.
| Practice · ROD | TTM Rev | Margin | Top Revenue Driver | Rev $ (12mo) | Top Expense Driver | Exp $ (12mo) | Combined $ |
|---|
1 · Driver classification — literature vs our data
2 · Multivariate combos — is 1 driver enough?
| Rank | Drivers | Adj R² | n obs |
|---|
3 · Segment-specific leading drivers
4 · EBITDA margin cross-section (n=52-59)
| Driver / ratio | r | p | Interpretation |
|---|
5 · Dashboard recommendations (empirically-grounded)
Four views, two axes
Each practice is scored on two axes (Revenue growth, EBITDA margin) using two independent data sources (Power BI financials — lagging but definitive; Dental Intel operations — leading and predictive). The 4-view grid surfaces divergences where leading signals hint at upside or downside not yet visible in financials.
Stages (both axes, both sources)
Foundation unstable — base low or margin fragile.
Momentum real, repeatability not yet achieved.
Operationally real — scaling systems, not proving demand.
Strong — gains come from efficiency & throughput.
Near ceiling — limited by capacity, not discipline.
Scoring weights (v2)
Revenue drivers vs expense drivers
Revenue S-Curve is scored on metrics that drive top-line: demand, capture, conversion, recall. EBITDA S-Curve is scored on expense drivers: cost ratios, labor waste, throughput efficiency. They move independently — a practice can be strong on one and weak on the other, and the levers for each are different.
Lever $ math — how every $ number is computed
Revenue levers flow through marginal margin:
EBITDA_impact = revenue_captured × marginal_margin × realization_factor marginal_margin = min(avg_margin + 25pp, 65%) realization_factor = 30% (12-month realistic capture) cap = 8% of annual revenue base (sanity) avg_margin uses REAL Gen4 TTM EBITDA margin when available
Expense levers flow 1:1 (reducing an expense ratio by 1pp on $10M revenue = $100K direct EBITDA):
EBITDA_impact = revenue_base × (current_ratio − target_ratio) × realization_factor target_ratio = portfolio top-quartile (p25 of expense-ratio distribution)
Correlation validation — empirical proof (24-mo within-practice)
24 months of DI history tested against grossProduction at lag 0 / +3 / +6 months. Within-practice correlation removes the practice-size confound and reveals whether changes in a metric predict changes in production.
How to read: within-practice correlations are much harder to surface than cross-practice because they reflect month-over-month volatility at one location. A |r| ≥ 0.30 at a forward lag is a real leading indicator; 0.10–0.30 is directional but noisy; <0.10 is not a statistically meaningful leading signal at this window. All levers still hold their math validity (deterministic benchmark gaps) regardless — this table tells you which signals deserve the most trust.
Data sources
v3 upgrade path
- Correlation validation — 24-mo DI history pulled; correlation against PBI forward production in progress
- Full INTACCT coverage — currently 84 practices with P&L decomp; needs full network
- Monthly refresh automation — currently snapshot; pipeline should run month-end
- Region/ROD rollup dashboards — mapping attached; rollup views next