Plan the staged path.
Loading the bound snapshot.
The path, by quarter
Each bar is a calendar quarter. 2026 includes November and December only. Sales, profit, and cash are sums. The ending pool is the last month of the period, not a sum.
Ending pool
The line is the ending cash pool. The gold mark is the planning pool from the snapshot.
| Period | Sales | Profit | Cash after tax | Ending pool | What is happening |
|---|
Capital rule
Three cash ideas sit on this page, and they are not the same number. Cash after tax is what the quarter made or spent. The ending pool is what is left after that, including the $660,000 the plan starts with. Peak need is the most cash the path has consumed at one moment. A quarter can show negative cash and still leave a positive pool.
The pool is modeled as $330,000 from a retirement rollover and $330,000 from other cash. That split is a hypothesis until a CPA and retirement-plan counsel sign it. Owner pay is booked once at Holding, $177,411 a year in this plan, and is already inside these figures.
Phase 1 has to fit inside the pool. Later purchases are not drawn from it. They wait on operating cash in 2028: the third cleaning route in May, car 4 in July, and cars 5 and 6 in September.
Each line, year by year
Ten operating lines, plus Holding. Holding is where owner pay is booked. It is not an eleventh business. Motors and Fleet is not listed here because it is only Foundry plus Obsidian added together. 2026 is November and December. Profit is after owner pay and shared costs. A line with no sales is still listed so it is not mistaken for a business that was dropped.
Fleet and supply
The path ends with six chauffeur cars, three Glean routes, and lawn and pool rigs. November 2026 does not buy that list. Phase 1 buys a bounded set. Everything after that waits on cash.
What the first wave puts on the road
From November 2026 through June 2027 the plan puts three luxury cars on the road for Obsidian: two GMC Yukon Denalis and one Mercedes S500 or BMW sedan. Foundry holds title and insurance and leases them to Obsidian. Glean starts with two commercial routes plus lawn, pool, and wash rigs. The third bought route is not in this wave. Novelle’s studio rent starts in January 2027. The other lines that do not need six cars on day one start without that garage.
The tightest point on this wave is $656,135 in June 2027, about $3,865 inside the $660,000 pool. Splitting the studio fit-out across two payments is what the notes use to describe a little more room, about $7,763. That is a planning note, not a signed lease.
What has to be true before the next purchase
The remaining $220,000 is not taken from the opening pool.
- Third Glean route, May 2028: hold the $40,000 down payment until operating cash that month is about $97,000.
- Car 4, July 2028: $60,000 only if operating cash is about $93,000.
- Cars 5 and 6, September 2028: $120,000 together only if operating cash is about $108,000.
Insurance and chauffeur coverage follow the cars that are actually on the road. Bind the first three before any paid trip. Treat cars 4 through 6 as a 2028 decision.
By December 2029 the model is describing six debt-free luxury cars, three routes, and the studio, if those tests passed. That is operating capacity at the end of the path. It is not a purchase order for November 2026. Vesper stays at zero until Obsidian sustains utilization or every deal is fully deposit-backed, so consignment inventory never hits the balance sheet.
Marketing
Marketing’s job is to fill the demand already inside each line. There is no separate marketing profit-and-loss. The sales column on the quarter sheet is what those motions are assumed to produce. This is not approval to spend.
Home services is one motion. Glean gets the household or the commercial account. Home Club turns that visit into a membership, which is what funds most of the shared desk. Simply Tech and Pitstop take the next job from the same household instead of buying the customer again.
Mobility is a second motion. Obsidian fills the cars that are actually on the road, first three, then more only if the 2028 cash tests pass. Vesper does not get its own ad budget on this plan.
Digital is a third motion. Lucent’s paid audit is the way into a retainer or the AI receptionist. Novelle adds creators only after letters of intent, and the studio fit-out waits on those letters.
| Line | Volume the path is built to reach | Sales in the last year | Still unproven |
|---|---|---|---|
| Glean | Commercial route accounts plus residential, lawn, and pool work | $2,913,952 | Local-services ad pilot |
| Home Club | 185 member households | $245,500 | Enrollment at the job, not a second ad buy |
| Simply Tech | About 5 referred installs a month | $139,832 | Partner licenses and insurance |
| Obsidian | Demand for the cars actually on the road | $2,500,802 | Search and hotel-concierge pilot |
| Lucent | 42 AI receptionist clients | $2,887,389 | Audit pilot before more ad spend |
| Novelle | Creator roster only after letters of intent | $1,256,514 | Signed creators before studio spend |
| Vesper | None until Obsidian utilization holds | $0 | Gated |
Before any of that spend is scaled
Every conversion rate in the plan is a hypothesis until a capped 30 to 60 day test reports cost per lead, bookings, and whether people stay. The caps already written down are $1,500 for Glean’s local-services ads and Home Club, $1,500 for Lucent’s audit test, $2,000 for Obsidian’s search and hotel-concierge test, and no studio spend for Novelle until two or three creator letters are signed. Measured results are not in hand.
People
Phase 1, November 2026 through June 2027, assumes a small team. Nobody on this list is hired. Counts and start dates are not in this sheet.
You stay in the operating seat through launch. Owner pay is booked once at Holding: $177,411 a year in this plan, covering salary, retirement, health, and employer payroll tax. A “share of owner pay” on a line is a management view. It is not a second paycheck.
Who the plan assumes
- One shared booking desk for Glean, Simply Tech, and Pitstop. Lucent’s AI receptionist is the after-hours overflow, not a second desk. The desk is modeled around $51,840 at the start and about $73,655 by the last year, funded by Home Club, homewatch, and a charge across the home brands.
- Employees for core cleaning routes and for peak chauffeur shifts, Thursday through Sunday, on the cars that are actually in service. Contractors take overflow cleaning and non-peak rides so payroll does not get ahead of booked work.
- Simply Tech uses licensed install partners, not a bench of in-house technicians. Insurance certificates come before any payout.
- Novelle is a lease and a creator roster, not a large day-one payroll.
The rule before the next person
Add an employee only when work already booked covers the fully loaded cost: wages, taxes, workers’ compensation, and the share of the bench. Do not add a person because a later year in the model looks profitable. Do not open a second booking desk. Workers’ compensation class codes, livery permits, and employment agreements have to exist before the first paycheck.
Chauffeur payroll is still an open record. The model uses hypothetical driver weeks. There are no actual timesheets, tips, commissions, or workers’ compensation bills. Collecting those records does not authorize offers.
Technology
Two offers are productized in this plan, and both hand work to the same booking desk. The year-by-year sales for each line are in the section above. Nothing here approves a software purchase or a vendor contract.
Lucent
The way in is a $495 audit of missed calls. If the buyer continues, that fee is credited toward a $1,500 monthly retainer or a $297 monthly AI receptionist that texts back and answers after hours. The plan’s unit view uses 42 receptionist clients by the last year and about $215 a month of margin per client after delivery cost. In the snapshot, Lucent’s own profit turns positive in November 2027, its cash turns positive in January 2028, and its tightest point is about $89,000 in December 2027.
A 45-day test with a $1,500 spending cap is what has to happen before those conversion rates are treated as real. The plan’s hope, a quarter of audits becoming retainers or subscriptions, is a hypothesis. There are no observed clients.
Simply Tech
The packaged job is a $2,600 whole-home Wi-Fi and wired camera install, done by a Florida-licensed low-voltage partner. Prism keeps 25%, about $605 a job, and carries no technician payroll on that path. The snapshot uses about five referred installs a month by the last year. Warm leads are expected from Glean visits and the shared desk, not from a separate ad budget. Profit in the snapshot turns positive in April 2027 and cash in May 2027. The tightest point is about $9,000 in April 2027.
Partner paperwork blocks payout: a Florida low-voltage license, a $1 million certificate of insurance naming Prism, and the 25% referral terms in writing. Whether every Wi-Fi scope needs a state license is still an open legal question.
The desk both of them use
Home services runs one human desk for Glean, Simply Tech, and Pitstop. Lucent sits on the overflow side of that desk, including nights. By the last year the plan uses 185 Home Club members, and their membership margin is what covers about two-thirds of the desk. Simply Tech jobs are quoted on that desk and paid only after the partner’s license and insurance are on file.
Legal and tax
This is the company map the cash plan assumes. It is not CPA, retirement-plan, or Florida tax advice. Nothing on this page is signed.
Who operates what
- Home services company: Glean, Simply Tech, Pitstop, and the shared desk, Home Club, and homewatch.
- Motors and fleet company: Foundry holds the cars, Obsidian drives them, Vesper is the gated sourcing line. The “Motors and Fleet” total is Foundry plus Obsidian only. It is not added on top of those two.
- Digital companies: Lucent and Novelle.
- Holding books owner pay and the consolidated tax view once. Property is the central facilities company. Neither is an extra operating line.
Intercompany paper has to exist before money moves. Foundry’s lease to Obsidian, the shared desk charge, and any marketing or rent between lines should say who owns the customer so a later sale can assign the contract.
Before the first check
- Confirm the companies, tax IDs, operating agreements, and separate bank accounts so retirement-plan money, owner cash, and customer receipts are not in one account.
- Document the $660,000 pool, modeled as $330,000 from a retirement rollover and $330,000 from other cash, before any vehicle is ordered. Qualification of that rollover is not established here.
- Prepare the Florida resale-certificate work and a master lease so Foundry can buy cars and lease them to Obsidian. Sales tax on that lease stream is a hypothesis until a CPA confirms it.
- Put assignment language in route contracts, shuttle retainers, the studio lease, and partner deals so a later sale is not blocked by a silent ban on assignment.
Still unsigned
- Written CPA and retirement-plan confirmation of the tax schedule, the Florida lease-tax treatment, and the $330,000 rollover.
- If a cleaning route is bought later, the seller’s records for 24 months, proof the accounts can be assigned, and the note terms. The first wave does not depend on that closing.
- Binding livery insurance, including airport coverage where a trip needs it, and permits in Palm Beach, Broward, and Miami-Dade before Obsidian takes a paid trip.
- Simply Tech partner agreements, licenses, and insurance before a referral is paid.
Still open
These items can change the cash path. They are not cleared because they appear in the workbook.