Skip to content
LuminaBusiness Advisory

The Review · a worked sample · invented records

What the finding looks like, before you pay for one.

This page shows the shape of one Review finding at full depth: the process map, one leak priced from records, and the argument for why it goes first. It exists because a firm that sells a $4,500 report should show you what a finding looks like — and because our clients’ real records are never published.

Read this first

Every figure on this page is invented. The company does not exist; its records were fabricated so the arithmetic can be shown in full. Every number on this site must be your input, a cited source, our own measurement, or a labeled assumption — this entire page is the fourth kind. It demonstrates the structure and reasoning of a Review; it is not evidence of client results.

The sample company

An invented HVAC-and-plumbing contractor. Twelve months of fabricated records.

Residential HVAC and plumbing, fourteen field technicians, three people on the phones. The records a real Review would read — the field-service platform, the phone logs, the books — were invented here at realistic proportions, then read the same way we read real ones.

Trailing-twelve-month revenue
$5.8M
Inquiries logged
260/mo
Estimates written
130/mo · avg $2,400
Estimates approved inside 14 days
34%
Estimates touched after day 14
9%
Open estimates older than 30 days
74 · $178,000
Maintenance memberships · lapsed this year without a call
610 · 118
Median days from job done to invoice sent
6

Invented records — the label applies to every figure above and below.

Document one · the process map

The chain, stage by stage — and where work stops moving.

Inquiry → Quote → Contract → Delivery → Invoice → Payment

  1. 01 · Inquiry

    3,120 logged in twelve months

    Phone logs show 214 more calls that rang out during peak weeks and were never logged anywhere.

  2. 02 · Quote

    1,560 estimates written — fast, usually same-day

    Writing the estimate is the last step anyone owns. What happens to it afterwards belongs to nobody.

  3. 03 · Contract

    530 approved inside 14 days

    The 1,030 that age past day 14 are touched 9% of the time. This is where the money stops moving — the finding below.

  4. 04 · Delivery

    560 jobs delivered

    Sold-to-scheduled runs clean here — eight days median. Not every stage is broken; the report says so when it is not.

  5. 05 · Invoice

    Median 6 days from job done to invoice sent

    Six days of quiet float on every job, and extras agreed on-site are on 31 invoices out of 60 sampled.

  6. 06 · Payment

    $96,000 older than 45 days

    Nobody owns the reminder. It goes out when the bookkeeper has a quiet Friday.

In a real report every stage carries its evidence: which records, which dates, which sample. Here the map’s job is to show the shape — one operation, read end to end.

Document two · the finding

Nine estimates in ten go quiet after day 14. Nobody follows up.

In this invented set: 1,030 estimates aged past day 14 in twelve months — 91% with no follow-up logged.

The owner’s opening guess was “we need more calls.” The records said otherwise: the leak sits after the estimate, not before the phone rings. The constraint is almost never the thing the owner names first — it is usually one layer deeper.

Estimates aging past day 14 each year
1,030
Of those, ever touched again — 9%
93
Close rate when chased after day 14 (their own records)
18%
Close rate when left untouched
3%
Untouched estimates × the gap between those two rates
937 × 15% ≈ 141 jobs
At the average estimate of $2,400
$338,000/yr

Raw estimate of the leak

$28,200/mo

The report never carries the raw number into the plan. A measured share of aged estimates is unrecoverable — duplicates, price-shoppers, dead addresses; in this invented set, 38%. What the plan carries: $14,000–$17,500 a month, with the haircut and its basis stated. Measured cost and estimated opportunity are kept apart, on paper.

And one more boundary the report states about itself: part of that fifteen-point gap is selection — the estimates someone chose to chase were likelier to close anyway. The gap suggests an opportunity; it does not establish the effect of follow-up. The pilot below is what tests how much of it is real. And one check comes before all of it: “no follow-up logged” is not “no follow-up happened” — the report verifies logging completeness against the phone records before it trusts the gap.

Document three · the priority rationale

Why this fix goes first — and what got sequenced behind it.

  • The money is already paid for.

    Marketing spent, estimate written, truck rolled. Recovering an aged estimate needs no new demand — which is why it outranks every idea that starts with “buy more leads.”

  • The fix lives inside tools already owned.

    A follow-up cadence with a named owner, built in the field-service platform the company already pays for. Process and configuration, not new software. And the build is recommended only if the conservative scenario — contribution margin, not revenue — covers its full cost inside the agreed payback period.

  • It unblocks what comes next.

    The same follow-up engine later carries membership renewals and the review ask. Building it first makes the second and third fixes cheaper.

What did not go first — and why

  • The 214 calls that rang out (≈ $9,000/mo on the same method): real, but smaller — and partly double-counted, because some callers tried again. Sequenced second.
  • The 118 lapsed memberships (≈ $4,300/mo): the renewal outreach rides on the follow-up engine from fix one existing. Sequenced third — order is arithmetic, not taste.

Document four · the change, specified

What we would change — and how we would verify it.

Trigger
An estimate is still open on day 14. The platform flags it to a named owner the same morning — not to a shared queue.
Action
The agreed contact sequence starts: a call, a text, one more call — spaced over ten days, written in the company’s voice, approved by the owner before the first one ever sends.
Exceptions
Declined in writing, already answered, duplicate estimate, or on the do-not-contact list — the sequence never fires. Nobody gets a robot follow-up after a human conversation.
Control
A flagged estimate untouched for 48 hours escalates to the service manager, by name. Failure is visible the week it happens, not at the quarter’s end.
Acceptance
No estimate leaves the pipeline unworked; every stop carries a reason on record; the chased-versus-untouched close rates are readable weekly, per person. The pilot’s inclusion list is fixed in advance, and every eligible estimate counts — contacted or not.
The decision after the pilot
Day 90: the same records, re-read against the day-one baseline — cohorts matched on estimate age, job type, amount and source, the same observation window, and costs with contribution margin read next to conversion. The lift is measured against that baseline, not assumed from the historic gap; without random assignment, some of the change may reflect timing or customer mix — the report says that too. Expand it, change it, or stop it: the records decide, and without a dated baseline we claim no credit.

The other documents

This sample is one finding. The report is nine documents.

  • 01The executive report
  • 02The leak, priced
  • 03The constraint map
  • 04Your six scores
  • 05The customer journey, mapped
  • 06The opportunity list, ranked
  • 07The automation verdict
  • 08The 90-day plan
  • 09Where to put the next dollar

What each one contains, what the two weeks look like, and the price — on the Review page.

What this sample cannot show

Three honest limits of an invented example.

  • Real records are messier — half-logged calls, renamed columns, a CRM nobody trusted for a year. Reading through that mess is most of the two weeks.

  • Your constraint may sit somewhere else entirely. This company’s leak was after the quote; yours may be in delivery, invoicing, or the phones. The map decides, not the template.

  • An invented company proves the depth of the work, not the size of your number. Nothing here is a promise of what a Review will find in your business.

The next step

Want this page written about your business — from real records?

That is what The Revenue Review is: two weeks, $4,500 fixed, and a straight answer — including “nothing here is worth fixing yet”, in writing.

Book the 20-minute call