Every point value, and what it feeds.

I — The two axes

Fit and capacity, scored separately, never averaged.

Fit asks whether Sabina would actually work in this shop: the trade, what the calls are like, who answers them after hours, and whether one person can say yes. Capacity asks whether the operation is at a size that could carry hiring her at all: call volume, front-office headcount, and what the phone already costs in labor.

Capacity is the one that filters. With no price published anywhere on Sabina’s surfaces, nothing else stands between an operation that is not ready yet and the queue — a strong fit score cannot carry a reading past a capacity floor. See chapter III.

II — The point tables

Nine questions, closed answer lists, whole numbers.

Fit sums to 10 from six answers. Capacity sums to 10 from three — two of them typed numbers, scored by range rather than a pill, because an owner knows a headcount or a call count as a figure, not a bracket.

This is the one weighted question, and the weighting is trade-first today: roofing carries two points because it is the trade her role-pack has been built against first, every other trade carries one, and a business outside the trades carries none. That is a statement about where the work is deepest right now — not that any sector is proven, and not that any is ruled out. Fit is six questions in all, and the five that are not the trade carry eight of its ten points, so an operation outside the trades can still read as a strong fit.

III — The verdict

Three bands, stated as numbers so they can be argued with.

capacity ≤ 3                          →  "early", regardless of fit
fit ≥ 7  AND  capacity ≥ 6             →  "strong"
fit ≥ 4                                →  "promising"
otherwise                              →  "early"

A low reading is never a bare no. Every band gets an honest sentence and a next step — the early band gets a standing-arrangement path, because the door stays open as the numbers move.

V — The leak calculator

One multiplier, four lines, nothing added that was not shown.

Chapters III and IV of the assessment are a second, separate tool: put in call volume, average job value, close rate and days to get paid, and it shows what unanswered calls and slow collections are costing — annualised, with every step of the arithmetic on the page.

calls  ×  25% unanswered      =  missed calls, monthly
missed  ×  close rate            =  lost jobs
lost jobs  ×  average job value  =  leak, monthly
leak, monthly  ×  12             =  leak, annualised

The 25% is the only multiplier in the calculation, and it is a published benchmark for missed inbound calls in the trades — not a measurement of any one reader’s phone. After-hours coverage changes which qualifier sentence renders under the figure; it is never read into the arithmetic above.

Cash tied up in accounts receivable is computed and shown separately, and is never added to the annualised figure — one is revenue that went to somebody else, the other is revenue that is coming but late. A stated share of the leak, set by the reader rather than picked for them, splits the annual figure into what is recovered and what is left on the table.

Two more published figures appear for context and carry no dollar value: average quote turnaround in this trade runs about 42hours, and speed itself matters because “The odds of qualifying a lead called in 5 minutes rather than 30 drop by 21 times.”