Where finance earns its value is not in assembling the numbers, it is in deciding what to do about them. These are working notes for CFOs, finance leaders, and FP&A practitioners: decision frameworks, pricing and margin analysis, cash-flow stress testing, and honest benchmarks of where AI helps the work and where it does not. Most posts ship with a free model or toolkit you can use the same day.
New here? Start with these. One framework, one free toolkit, and one benchmark, together they cover how I think about strategic FP&A, how I package it into tools you can run, and how I test where AI actually adds value.
The three-layer model that moves analysis past "what happened" and into "what should we do," with a worked manufacturing example.
Read the framework →Assembly eats the time budget and the interpretation gets cut. A free toolkit builds the deck, dashboard, Excel, and commentary from raw data.
Read & get the toolkit →Opus 4.8 vs Sonnet 4.6 tested across effort levels on a real close task: five conditions, seven scoring dimensions, one clear risk warning.
Read the benchmark →The core thesis of this site: finance creates value when it shapes the next decision, not when it describes the last quarter.
Three layers, Reflection, Diagnosis, Decision Modeling, with a worked example and links to every supporting model.
Read →What actually breaks when a company runs without an analyst, and why serious competitors can't afford to find out.
Read →3 weeks cut forecast error from 12% to 7% and not one decision changed. Find the threshold before you spend the time.
Read →40 variances, 40 comments, and a CFO who learns nothing. Two filters turn that into 3 signals worth acting on.
Read →Tested, not hyped. What changes when one analyst has serious leverage, and what the tools still cannot do.
A practitioner benchmark across five conditions and seven scoring dimensions, with the full research paper included.
Read →AI didn't replace the analyst, it removed the ceiling on what one skilled analyst can do. What that means in practice.
Read →Each of these pairs a specific business question with a free, working model, on margins, pricing power, and cash runway.
A free toolkit that turns raw data into a full board package, deck, dashboard, Excel, and commentary.
Read →21% higher contribution with fewer units shipped, using elasticity to find where pricing power actually exists.
Read →By the time you feel a cash problem it's usually too late. A 26-week stress test that surfaces it early.
Read →Sales rose 18% and gross profit fell. The discount needed 60%. One formula that settles it before you approve the promotion.
Read →Don't ask how much volume a price rise will cost you. Ask how much you can afford to lose. A one-minute answer that ends the argument.
Read →A customer at 3.5 times the revenue returning one eighth of the contribution. Gross margin measures the product, not the customer.
Read →$50,000 more revenue and $15,000 less contribution, on the same factory. When capacity binds, the star product is often the worst use of it.
Read →Cost plus 40% gave $84. A rival charged $130 for the same function and outsold them. Cost is the floor, not the anchor.
Read →The same downturn modelled both ways: $825,000 of profit as a sensitivity, $556,000 once the linked drivers move together.
Read →Every model referenced in these posts lives on the Tools page, ready to download and run on your own numbers.
Browse Tools