The TAG method
See the Truth. Implement Actions. Achieve your Goals. The tagline spells the name — the method is a picture.
Method
A firm I ran was break-even when I arrived — a real business with twenty people, revenue coming in, and a bottom line that oscillated between minus five and plus five percent around an average of zero. Nobody could say why it moved, because nobody could answer the question "which parts of this make money?" Nine months later it held sustained double-digit net margins, with peak months above 25%. Every part of that turnaround is on this diagram: fix the left endpoint, fix the right endpoint, then draw the line.
Most businesses run the picture in the wrong order. The goal gets set first — a revenue target, a margin ambition, a number for the year — before anyone knows what is actually true about the P&L, the unit economics, or the funnel. That goal is fiction: a point floating on the page with no line to it, because a line needs two real ends. TAG refuses to draw the line until both endpoints are real.
Truth — the left endpoint
Rebuild the numbers until the real economics are visible. Not the numbers the accounting system happens to produce — the ones that answer operating questions. Revenue per payroll dollar. Marketing spend to booked revenue. What each service line actually costs to deliver. Where the intake funnel leaks. In the firm above, this meant designing and building a business-intelligence system from scratch: acquisition, production, billing, reconciliation, forecasting, per-person scorecards. Truth is not a feeling about the business. It is instrumentation.
Goal — the right endpoint
Fix the goal the truth can support. Once the real economics are visible, most inherited targets dissolve — some were never reachable, others were too timid. In that firm the truth supported a specific, dated commitment: durable double-digit net margins within a year, without cutting into the capacity that produced the revenue. A goal set this way is not an ambition. It is a destination with known coordinates, which is the only kind a line can be drawn to.
Action — the line between them
Actions come last, because an action is a bridge and you cannot draw a bridge until both banks are fixed. With both endpoints real, the line almost draws itself: the work is whatever carries the business from left to right, and nothing else. In that engagement the line ran through four interventions — staffing and incentive-compensation redesign, so pay moved with measured contribution; pricing; collections discipline, held near 100% with chargebacks driven to zero; and an intake rebuild that took the consult show-up rate from 24% to 90%. None of those were generic best practices. Each was a span of the same bridge.
Achieving the goal is not a further step — it is what standing at the right end of the line looks like. In that firm: break-even to sustained double-digit net margins with peak months above 25%, production labor ROI from 4.7X to 6.6X, marketing ROI from 4.4X to 7.5X, average sale value more than doubled — each confirmed in a signed reference letter from the owner, because by then the instrumentation existed to confirm them.
Why the picture holds up
I spent fifteen years practicing nuclear medicine before I ran businesses. Reading scans is fifteen years of daily discipline at fixing endpoints: establish where the patient actually is before deciding where treatment should take them, and only then choose the intervention that connects the two. TAG is that clinical habit applied to a P&L. It was not invented for marketing; it is how I was trained to think, and I have run it inside companies I have led and with the owner-operators I have advised.
The other thing the clinical years explain: I build the instrumentation myself. Most executives at this level can specify a reporting system and wait two years for IT. I design and ship it with AI-assisted development — four production applications so far — which means a 10–50 person firm gets enterprise-grade visibility at small-company cost, live within a quarter. Judgment and implementation in the same person is the scarce thing. Analysis, these days, is nearly free.
Publishing October 2026
A working demonstration of this instrumentation — a full operating dashboard built on entirely synthetic data — publishes here in October. Until then, the capability is described above and shown in conversation.