The whole story, on one page

If you read nothing else, read this page. Everything after it is detail.

• What this is: a small family of online businesses, run day-to-day by Eva - an artificial-intelligence helper - under Glen's ownership. Permit Intel came first; the excitement this week is Upwork and Lode Star.
• The scoreboard today: $0 made, 0 paying customers. Everything is built and being tested. Nothing is income yet, and this report never pretends otherwise.
• The ask: $30,000 from Mom and Dad, as a gift. No loan, no repayment, no ownership share.
• How the money is protected: it moves in three stages - about $5,000, then $10,000, then $15,000 - and each stage unlocks only after paying customers exist. If the tests fail, the later stages are never spent.
• How Glen lives while this ramps: Glen draws $700 a week from this same $30,000 for the family while the businesses get going. At that rate the fund alone lasts about ten months - which is exactly why the spending is staged and the checkpoints matter: the businesses have to start earning before the draw runs the fund down.
• What it costs to run: about $240 a month, for everything.
• The exciting part: Upwork sent 8 proposals on its first day - one already viewed - and Glen's written prediction is that our first paying customer comes from there. Lode Star watches government demand for language services, and Eva's translation has already passed a graded test with zero critical errors.
• The dates that decide: November 12 for Permit Intel, November 23-25 for the other readouts. The pass and fail numbers were written down in advance - they decide, not hope.
• The worst case: the fund gets used up - Glen's weekly draw plus any stage deployed behind evidence - in about ten months if nothing earns. What can never happen: a family debt. No one owes anyone anything beyond this gift.
• How you'll hear: a short, plain update from us after every checkpoint, good news or bad - starting with November 12.
• And an invitation: let's sit down and talk it through. No deadline, and no pressure.

Bottom line: built, being tested, $0 so far - $30,000 as a gift, staged behind customers and drawn on at $700 a week, first verdict November 12.

A family of businesses:
one scoreboard,
kept honest

What we are building, why the money is safer here than it was in the cleaning business, what it costs, what it could earn - and the honest scoreboard, which today still reads $0. Permit Intel came first. The excitement this week is Upwork and Lode Star.

Where things stand today - plainly

• Sales so far: $0. Customers so far: 0. This is a business in testing, not a business with revenue.
• The test: a disciplined 21-day market test. First outreach goes out October 22-23, 2026. We set the pass and fail lines in advance, and we get a clear verdict on November 12, 2026.
• The risk if it fails: under $200 total. No lease, no equipment, no employees, no debt.
• The goal it serves: $5,000 a month. At $149 a month, that takes 34 subscribers - a number this report shows its work on.
• The ask in this report: $30,000 - the amount you were ready to provide for Growth Cleaning - for the whole family of businesses Eva now runs, released in stages that only paying customers unlock. Page 2 brings this report up to date as of October 28, and the last sections show exactly where every dollar is planned to go.
• How Glen lives while this ramps: Glen draws $700 a week from this same $30,000 for the family while the businesses get going. At that rate the fund alone lasts about ten months - which is exactly why the spending is staged and the checkpoints matter: the businesses have to start earning before the draw runs the fund down.
• What the $30,000 is for, in three lines: about $5,000 to expand what passes its test and set the business up properly; about $10,000 for marketing we can measure, spent only after customers exist; about $15,000 held in reserve so the work can grow without pressure. Each stage unlocks only behind paying customers - the full case is on page 4.

Prepared by Glen, with Eva - my AI operations partner, who runs the day-to-day work of the business under my ownership and direction.
Sioux Falls, South Dakota · October 21, 2026

How to read this: Short on time? Read pages 1-3 - the whole ask is on page 1. Want the full story? Read straight through; each part answers one question, and the questions are the headings. One word we use a lot: a checkpoint is a date with a number we agreed in advance - if the number is not met, the next step does not happen.

What’s in this report

• Key points - the whole story on one page - page 1
• The short version - this page
• What is new since last week - page 3
• What are you asking us for? - page 4
• What happens if it doesn't work? - page 5
• How will we know if it works? - page 6
• What have you actually built? - page 7
• What did it cost to build? - page 8
• Who checks the numbers? - page 9
• How did you find this business? - page 10
• Has anyone done this before? - page 11
• Who else already does this? - page 12
• What does the money math look like? - page 13
• What could the next year look like? - pages 14-16
• What does it cost to run - and what will you need later? - page 17
• Why not the cleaning business? - page 18
• What could go wrong? - page 19
• One last thing - and an invitation to talk - page 20

Bottom line: the business is built and being tested, it has made $0 so far, and we are asking for $30,000 - staged behind paying customers, with Glen's $700-a-week living draw coming from the same fund, stated openly.

Update - October 28, 2026

In one sentence: in the week since this report was written, the work has grown from one business into a small family of them - and the honest scoreboard has not moved yet.

Who is Eva?

Eva is a computer program - an artificial intelligence - who works for our business. All day, she does the office work: the research, the writing, the bookkeeping, and watching for replies. She cannot spend money: Glen approves every dollar himself. She cannot sign anything. And she is not allowed to say something is sold until the payment record proves it. Glen checks her work, and a second system (page 9) re-checks the numbers every night.

The exciting part first: Upwork and Lode Star

Upwork - Eva's storefront on the world's biggest marketplace for freelance work - went live this week with all six of her service lines. Glen set up the bank and tax side himself, the account is verified, and 8 proposals went out on the first day; one has already been viewed. Glen's prediction is written down: our first paying customer comes from Upwork.

Lode Star is the newest door: it watches what governments are about to buy in language services - translation, transcription, and captioning - so Eva can offer her work at exactly the right moment. Her translation has already passed a graded test with zero critical errors, and the watch runs every day. Plainly: no customers from Lode Star yet - it is a watch, not a business yet - but it is the door we are most curious about.

What we have done

• Permit Intel is live, and easier to try. The website and the $149-a-month signup are live, and we added a 14-day free trial with no credit card required - both big competitors demand a card up front, and we would rather remove that hurdle.
• The first outreach went out - carefully. A small paid ad test ran and was stopped at $58.26 after 27 clicks, with one real sample request from a builder, Buck Up Construction. A wave of contact-form outreach brought one human reply, from Carol Smith. Small numbers, honestly reported.
• The main email launch is deliberately on hold. Our email score sits at about 70. That score - email providers call it deliverability - is how much they trust our mail; below 80, too much of it lands in spam folders. The bar we set for ourselves is 80 before the big wave sends. One sending domain's signature problem was found, fixed, and proven fixed. We will not spam our way into a reputation problem - the wave launches when the score earns it, not before.
• The checking system now re-checks everything, nightly. A checking system called Ground Truth (page 9) now re-verifies the whole operation every night. Its first full sweep, on October 28, covered all 10 areas it watches - 10 out of 10.
• Eva can now work on Glen's own laptop - with safeguards, and not yet fully proven. Glen approves any payment himself, and some surfaces are hard-blocked. Plainly: Eva is still being tested against a written driving exam, and she has not passed it yet.
• Two more storefronts opened. Contra offers Eva-run services at $49 to $195; it is live and can take payment. On Upwork, the account is verified, Glen set up the bank and tax details himself, the profile now covers all six service lines, and 8 proposals went out on the first day, October 28 - one has already been viewed. Glen's registered prediction, written down in advance: the first paying customer comes from Upwork.
• Two small test businesses are in the field. F5 sent 20 claim-verification offers to newsletter writers on October 27. Lode Star watches real government demand for language services. So far: 0 customers from either. They are tests, not results.
• Costs grew modestly. New since October 21: a rented work server at €29.91 a month (about $33) on a one-month trial, an Upwork membership at about $19.99 a month plus proposal credits, and a one-time $11.08 domain for Contra. Rebuilt from its pieces - $100 for the Muse subscription Eva runs on, $69 for the email platform, about $33 for the server, about $20 for Upwork, about $15 for domains and mailboxes - the whole operation now runs at about $237 a month. Call it roughly $240; that figure is approximate.

What we are planning

• November 8: a review of the whole family of businesses - what stays, what stops.
• November 12: the Permit Intel verdict, on the pass and fail lines written in this report (page 6). A second city - San Diego County - is considered only if Austin passes.
• The held email wave launches only when our email trust score reaches 80. No date is promised; the checkpoint decides.
• Upwork gets its first review after 10 proposals. The walk-away line is set in advance: 20 proposals with no real interest, and that line closes.
• Readouts already on the calendar: the rented server's one-month verdict on November 23, F5's results on November 24, and Contra's keep-or-stop line on November 25.

One thing in this report has changed size: the ask itself. It is no longer $15,000 for Permit Intel alone. It is now the $30,000 you had ready for Growth Cleaning, for the whole operation above, released in stages that only paying customers unlock. The full case is on page 4, and the ask figures throughout this report have been updated to match.

Bottom line: a great deal got built in a week, the scoreboard is still $0, and every date that decides what happens next is written down.

What are you asking us for?

In one sentence: this operation does not need the money to exist - the $30,000 would let it grow faster and last longer, and it is released only as paying customers unlock each stage.

Let me say the headline honestly first: this business is already built and testing for under $200. Nobody needs to write a check for it to exist. This capital is not for survival. It is for acceleration and staying power - growing faster, with the breathing room to do it properly.

The ask: $30,000 in startup capital - the full amount you were ready to provide for Growth Cleaning. The difference is what it buys and how it is released. It now funds the whole operation described in this report - Permit Intel, the Contra and Upwork storefronts, the small test businesses, and the systems underneath them - and it is released in three stages, each unlocked only by paying customers from the stage before. If the customers never come, the later stages are never spent.

What kind of money is it?

A gift - plainly. A gift from Mom and Dad, freely given: not a family loan, and not an investment with a return attached. There is no repayment schedule, no interest, and no ownership share - nothing is owed back in money. The staging, the stopping lines, Glen's personal approval of every dollar, and the update after every checkpoint are not legal terms, and nobody is imposing them. They are Glen's promise about how he will treat this gift: the checkpoints protect the gift's purpose even though nothing legally forces them to. What Mom and Dad receive is the reporting: a plain-language update after every checkpoint (page 5), a simple accounting of every stage as it deploys, and the standing invitation to ask anything at any time. The money is held in the business account, separate from household money; if a checkpoint kills the plan, the undeployed stages are simply never spent, and the unspent balance stays in that account - visible in every update. Glen's $700-a-week living draw comes from this same fund, and it is stated openly wherever this money is discussed - it is part of the accounting in every update.

How Glen lives while this ramps. Glen draws $700 a week from this same fund for our family of five - stated plainly, not tucked away. At that rate, the fund alone lasts about ten months, which is exactly why the spending is staged and the checkpoints matter: the businesses have to start earning before the draw runs the fund down. The draw is counted in the accounting of every update, right alongside the stages. And every business the fund supports still carries a kill line - each business in the October 28 update (page 3) carries its own stop date or stop number, written in advance.

Why not Growth Cleaning?

You were ready to back me with $30,000 for Growth Cleaning. I stopped it - that call was mine, and I would make it again - because that business would have spent your money up front: equipment, supplies, and my own hours every single day, with the whole amount at risk from the first week. Here is the same $30,000, placed where almost none of it moves until customers exist first. Before deciding, I researched that program properly - its own materials, its reviews, and what experienced cleaning operators say. The honest findings are on page 18. The short version: the challenges are not the cleaning. They are the people, the cash timing, and a price you only learn on a sales call.

Where the $30,000 is planned to go - in stages

Planned use	Amount
Stage 1 - after a verdict with paying customers: expansion and setup - additional city launches for Permit Intel (setup, domains, verified prospect lists), plus business legal setup and counsel review	$5,000
Stage 2 - unlocked by paying customers from Stage 1: measured marketing across the businesses that are earning - spent only where returns can be measured	$10,000
Stage 3 - unlocked the same way: operating runway and reserve - covers the roughly $240-a-month running costs (updated October 28) for years, and lets the operation build subscriptions without income pressure	$15,000
Total startup capital requested	$30,000

What the capital buys is speed and staying power, not survival: more cities sooner, real marketing tests instead of guesswork, and room for subscriptions to build month over month. A business needing only 34 subscribers across a few cities to reach $5,000 a month, at roughly 93% gross margin over running costs, is a reasonable place to buy those things.

Bottom line: $30,000, as a gift, in three stages - and no stage moves until paying customers unlock it.

What happens if it doesn't work?

In one sentence: the worst case is small, staged, and survivable - and you will hear about it from us, on a schedule, good news or bad.

The worst case, in plain words

If this fails, what is actually lost is only the money that was deployed behind evidence - one stage at a time, never the whole $30,000 at once. The later stages never deploy after a kill verdict: if November 12 goes against Permit Intel, the growth stages stay shut, and the unspent money stays unspent, in the business account, accounted for openly. What is at risk is the fund itself: Glen's $700-a-week draw, plus any stage deployed behind evidence - at that draw, the fund alone lasts about ten months if nothing earns, and the checkpoints exist so a failing business is stopped, and faced, while money still remains. What is not at risk: anyone's obligation. No family member owes anyone anything beyond this gift, and a business failure never becomes a family debt - that is a commitment, not a hope.

How you will be kept informed

After every major checkpoint, you get a short, plain-language update - the first one is the November 12 verdict - good news or bad, with a simple accounting of what has been deployed at each stage and what remains. Investors hear from us on a schedule, not when they ask.

If Glen were unavailable

The records of this business live in a system that keeps its own books and verifies them every night (page 9). Glen personally approves every dollar that moves. If Glen were unavailable, the money simply stops moving - nothing in this operation spends itself, and nothing can be quietly spent by anyone else, including Eva.

Bottom line: the fund is protected by staging and checkpoints - the draw and the stages are watched against the calendar together - a failure never becomes a family debt, and you will hear every verdict from us, on schedule.

How will we know if it works?

In one sentence: we wrote down exactly what success and failure look like before we started, and November 12 is the day we find out.

The point of a small test is that it can actually fail - and that we agree in advance what failure looks like, so nobody moves the goalposts afterward, including me.

How the outreach works

• 294 Austin prospects in total, identified and verified by hand-checked research.
• October 22-23: first contact through website contact forms, starting with 43 businesses that have a usable form.
• Email follows as the sending address completes a standard anti-spam warm-up - a normal process where a new email address gradually proves itself trustworthy to Gmail and Outlook. Its progress is reported daily.
• Anyone who replies asking for a sample gets a real sample brief for their trade, so they judge the actual product, not a sales pitch.

Timeline

Bottom line: the test costs under $200, the pass and fail lines were written down before it started, and November 12 gives the answer.

What have you actually built?

In one sentence: we check the city’s new permits every weekday morning so small contractors don’t have to, and send each one only the jobs that fit his trade and his area.

Every day, the City of Austin publishes the building permits it has issued - who is building what, where, and for how much. It is public record, and the city's collection holds about 2.38 million permits, updated daily. Anyone may look at it for free.

Almost no contractor does. The city's website is built for filing permits, not for finding work. Searching it properly takes hours a week, and most small companies would rather be on a job site.

Permit Intel does that searching for them. Each weekday at 7 a.m., every subscriber gets one email, ranked so the best leads are first, plus a simple spreadsheet file they can open, sort, and work from. It covers only their trade and only the ZIP codes they serve. No account to log into, no software to learn - because people who work in the field do not sit at a desk checking screens.

Who it is for

Not the big builders who pull the permits. Permit Intel is for the companies who come after them - what we call the adjacent trades:

• Fence and deck companies
• Landscaping and concrete companies
• Solar and gutter companies
• Final-clean and roll-off dumpster companies

Why a permit is a buying signal

• A new pool permit means a fence will be needed within weeks - pools require fencing.
• A reroof permit is the moment to talk to that project about gutters and solar panels.
• A new build will need a roll-off dumpster, concrete work, and a final clean before anyone moves in.

The large national services mostly sell raw leads to the big trades. Permit Intel sells these timing signals to the trades that follow the big jobs - a corner of the market nobody else packages.

Price: $149 a month, month to month, cancel anytime. No long contract.

Bottom line: for $149 a month, a small contractor gets the right jobs in his inbox by 7 AM - and never has to search the city's website again.

What did it cost to build?

In one sentence: the whole business cost $199.16 to build. It ran at about $180 a month then; at today's bigger size it is about $240 a month (page 17) - and two subscribers still cover it.

What was spent or committed	Amount
Website and domains (including the trypermitintel.com domain at $11.08, and business email hosting at $36 a year)	$47.08
Sending domains and mailboxes (separate addresses used for outreach, so customer mail stays clean)	$83.08
Lemlist sending platform - first month (the tool that sends and tracks outreach email)	$69.00
Total spent or committed to date	$199.16

What it costs to keep running

• About $240 a month, all in, at today's size. It was $180 a month when this business stood alone; the itemized breakdown of today's figure is on page 17.
• Payment fees, only when sales exist: Stripe, the payment processor, takes 2.9% plus 30 cents per charge.
• Nothing else. No employees. No office. No inventory. No vehicle.

That low cost base is not a side detail - it is the strategy. A business that needs 200 customers to survive has to grow fast or die. A business that breaks even at two customers can grow at the speed the evidence allows.

Bottom line: the whole business cost $199.16 to build, and two subscribers cover what it costs to run.

Who checks the numbers?

In one sentence: we built a record system that verifies itself against the bank and website records every night, answers our questions only from those records, and says “I don’t know” instead of guessing - so any number we report to you can be traced back to where it came from.

Why it exists. On the morning of October 22, a glitch in our monitoring reported that Permit Intel had its first paying customer. It would have been easy - and wrong - to celebrate. Instead we checked three separate ways: the bank-side payment records, the website’s own records, and the mailbox the report pointed to. The customer did not exist. We caught it and corrected it within the hour, and then we spent the rest of that day building the system on this page, so that from now on good news has to be verified before it is believed - by us, and by you.

Ground Truth. The system is called Ground Truth. It keeps the business’s records in one place - every permit report sent, every dollar, every customer, every message we rely on. Every night it checks its own numbers against the actual payment records and website records, and it test-restores its own backup to prove the backup really works. Then it reports one line: books balance, backup restores, feed normal. If anything disagrees, it does not smooth it over - it says so, in plain terms, the same morning.

The Oracle. On top of Ground Truth sits a question-answerer we call the Oracle. We ask it plain questions - “Why was this company in Tuesday’s report?” “How many subscribers do we have?” “How accurate is this source of ours?” - and it answers only from the records, showing where each answer came from. When the records do not hold the answer, it says so instead of making something up. It even keeps a running score of our own information sources: the monitor that raised the false alarm now stands at 50% in its books, and we treat its reports accordingly.

It notices things, too. Each morning the Oracle also points out the single most interesting pattern it found in the week’s permits. The first morning, it noticed that one Austin contractor was named on 91 of the city’s 973 new permits that week - nearly one in ten. Nobody asked it to look. That kind of early warning is exactly what our future subscribers will be paying for.

What it costs, and what it doesn’t change. Eva built Ground Truth and the Oracle in one day, October 22, on the systems we already pay for - they added nothing to the $180 monthly running cost in this report. The system even survived the computer it runs on being replaced twice that same day, because its backups are kept separately and restored cleanly both times. None of this changes the test, the November 12 decision, or the ask. It changes how much trust the numbers in this report deserve: every one of them can now be traced back to a record, any time you ask.

Bottom line: nobody here grades their own homework - a separate system re-checks every number, every night.

How did you find this business?

In one sentence: I went looking for a business that runs on systems instead of my hours, and this is the one that fit.

In October 2026, Eva ran a structured search for online businesses I could own - businesses that did not need me on a ladder, in a truck, or managing a crew. Each idea was scored on a simple question: can this actually run by automation, day after day?

Permit-data subscriptions won that search, for three practical reasons:

• The product makes itself. The deliverable is data, pulled from the city's free feed and re-sorted automatically. There is nothing to manufacture, stock, or ship.
• The whole operation automates. Billing, sign-up, and the daily delivery can all run as a system, with Eva operating it and me making the ownership decisions.
• Other people have paid for this for decades. This was not a guess about a brand-new habit. Two long-running companies, described on the next page, prove the demand is real and durable.

I approved a small, capped test on October 20, 2026. Within days, the website, the payment system, the subscriber pipeline, and the first list of Austin prospects were built - and rehearsed end to end, with test briefs produced for real trades, before a single real customer was contacted. That rehearsal-first habit is deliberate: we would rather find the cracks ourselves than have a paying customer find them.

Why Austin first

Austin publishes its permit data freely and keeps it current, it has a large and active construction market, and one city is the right size for a first test. If the model works in Austin, the same system can be pointed at the next city. If it does not work, we will know it cheaply and clearly.

Bottom line: we went looking for a business Glen could own without working in it every day - this is the one that survived the research.

Has anyone done this before?

In one sentence: people have been paying for permit information since 1989 - this is a proven kind of business, not a new fad.

This is not a new idea, and that is a point in its favor. Selling building-permit information is a decades-old, durable business that began long before the internet made it easy.

• 1989 - Construction Monitor. Founded by David B. Mineer Sr. after watching tradespeople stand in line at permit offices to learn what was being built. The original product was exactly what it sounds like: collect the permit records and mail them to builders. Today it is an online subscription covering more than 70 areas, it employed about 50 people as of 2018, and its data is licensed through Bloomberg.
• 1992 - HBWeekly began selling permit leads, and is still in that business.
• 2022 onward - the modern wave. A company called Shovels, founded in 2022, sells permit data at $599 a month (Basic) and $999 a month (Pro), and has raised $6.5 million from investors. Another company, PermitFlow, raised a $54 million funding round in December 2025.
• 2025-26 - small entrants. A wave of smaller services has appeared at prices much closer to ours: Permitlify at $79 / $149 / $349 a month, PermitMap at $79-$299, PermitPing at $179 a month per trade, for each city area, ZoneNotice at $99, and Lone Star Leads at $99 / $150 in Texas.

What changed - and why there is an opening now

For most of that history, the hard part was collecting the permit records - driving to offices, copying filings, mailing lists. That work justified high prices.

Cities now publish the same records free, online, every day. So a small operator's cost for the raw material is close to zero. The value today is not collection. It is curation: sorting the flood by trade and neighborhood, ranking what matters, and delivering it at the moment it is useful. That shift is what lets a business run from a laptop compete on usefulness instead of on headcount.

Bottom line: selling permit information is a 35-year-old business, not a new fad - we are doing it smaller, cheaper, and for the little guys.

Who else already does this?

In one sentence: the big companies sell everyone the same raw lists at high prices; we win by being local, specific, and cheaper to run.

The competition falls into two groups: the nationals (Construction Monitor, Shovels) - broad, expensive, and aimed at large firms and data buyers - and the new small entrants selling raw permit leads, mostly to the primary trades who pull the permits themselves.

Our four advantages

• 1. The follow-on signals nobody else packages. We connect one trade's permit to another trade's opportunity - pool to fence, roof to gutters and solar, new build to dumpster, concrete, and cleaning. That cross-trade view is the product, not a feature.
• 2. Email only - no login, no app to check. Contractors in the field do not log into portals. A ranked email and a spreadsheet at 7 a.m. fits how they actually work; a fancy platform does not.
• 3. One city area, done deeply. Every Austin trade and ZIP code, with a prospect list checked by hand, at $149 a month against $599 and up for the nationals.
• 4. A cost base with no staff. Eva operates the service, so the business runs on software costs instead of payroll. It can be profitable with a handful of subscribers, which means it can afford to be patient and personal in a way a 50-person company cannot.

Bottom line: others do this too; our edge is small crews, a fair price, and a checking system that admits mistakes.

What does the money math look like?

In one sentence: at $149 a month each, the math stays simple - 5 subscribers is $745 a month, and 34 is the $5,000 goal.

Everything below is arithmetic at $149 a month per subscriber. These are scenarios - what the business earns if it reaches each number - not a forecast that it will.

Subscribers	Monthly revenue	What it means
5	$745	Proof that strangers will pay
10	$1,490	A real, if modest, income stream
20	$2,980	More than half the family goal
34	$5,066	The $5,000-a-month goal
50	$7,450	Requires more than one city (below)

The honest pacing note

We have identified and verified 294 Austin businesses as the reachable market - 139 we can email, and 155 reachable by website form or phone. Thirty-four subscribers would be about 12% of that entire list. That is ambitious for one city, and we will not pretend otherwise.

So the growth path is replication: the same system, launched city by city. Each new city area costs roughly $100 in setup plus list-building time, because the permit data is free in city after city. The goal is not 34 subscribers in Austin. It is a handful in each of several cities.

One more unknown, stated plainly: we do not yet know how long subscribers stay (called churn). The November test measures demand - replies and sample requests - not retention. Retention is the next question after this one.

Bottom line: the math is simple on purpose - one price, $149, and every size of business is just that price times the number of subscribers.

What could the next year look like?

In one sentence: if things go the middle way, this passes $5,000 a month next June - and every path on these two pages depends on the November 12 test first.

This is the full projection, month by month - the same three paths, with nothing rounded away. Where things stand today, stated plainly: 0 subscribers and $0 in revenue. The price is $149 a month, and it stays $149. Running costs are about $180 a month, so two subscribers cover the costs. The family goal - $5,000 a month - takes 34 subscribers ($5,066).

One condition before anything else on these two pages: if the November 12 verdict is that the test failed, this projection is void. We stop, and none of these paths happens.

Month	Slower - subs	Slower - $/mo	Middle - subs	Middle - $/mo	Faster - subs	Faster - $/mo
Nov 2026	2	$298	3	$447	5	$745
Dec 2026	3	$447	6	$894	11	$1,639
Jan 2027	5	$745	10	$1,490	19	$2,831
Feb 2027	6	$894	14	$2,086	28	$4,172
Mar 2027	8	$1,192	19	$2,831	39	$5,811
Apr 2027	10	$1,490	25	$3,725	52	$7,748
May 2027	12	$1,788	31	$4,619	66	$9,834
Jun 2027	14	$2,086	38	$5,662	82	$12,218
Jul 2027	16	$2,384	45	$6,705	99	$14,751
Aug 2027	18	$2,682	52	$7,748	117	$17,433
Sep 2027	21	$3,129	60	$8,940	136	$20,264
Oct 2027	24	$3,576	68	$10,132	156	$23,244

The shaded rows mark the month a path crosses $5,000 a month: the faster path in March 2027 (39 subscribers, $5,811) and the middle path in June 2027 (38 subscribers, $5,662). The slower path does not cross within this year.

Bottom line: these are three honest paths, not a promise - even the middle one takes until the second year to reach the goal.

What do the three paths mean?

In one sentence: the slower path still pays its own way, the middle path is the plan, and the faster path would mean bringing in help.

The slower path - about $20,700 collected in year one

Growth is real but thin. The year ends at 24 subscribers and $3,576 a month. The $5,000 goal is not reached in year one - but the business covers its own costs almost from the start, because that only takes two subscribers.

The middle path - about $55,300 collected in year one

The path we would plan around. A handful of subscribers by Christmas, steady growth in Austin through the spring, a second city in April, a third in the fall. It crosses $5,000 a month in June 2027 and ends the year at 68 subscribers, about $10,100 a month.

The faster path - about $120,700 collected in year one

The test lands hard, word spreads, and a second city opens in March instead of April. It crosses $5,000 a month in March 2027 and ends the year at 156 subscribers, about $23,200 a month. Honestly, this path creates its own problem: serving 156 subscribers well would outgrow one operator, and the help or extra automation that takes is a real cost the numbers above do not include.

How the year is spent

Rest of 2026: prove it in Austin. Early 2027: deepen Austin - the suburbs around it, being found on Google, and word of mouth through trade suppliers. Spring 2027: a second city, same system, same price. Fall 2027: a third city, if the first two are carrying themselves. This is the year the $30,000 described later in this report is meant to fund - the staged expansion, the marketing, and the staying power to reach them.

What these paths assume

The price stays $149. Of the contractors who try the brief free for two weeks, about 1 in 4 (slower path), 1 in 3 (middle), or 2 in 5 (faster) end up paying. Some subscribers cancel along the way: in a typical month, about 8 of every 100 on the slower path, 6 of every 100 on the middle, 4 to 5 of every 100 on the faster - and every number on the previous page already has those cancellations subtracted. Growth comes from outreach, search, and suppliers; no paid advertising is assumed.

What could knock it off track

Three things, plainly. The test itself - November 12 decides whether any of this starts. The city's permit feed - the brief can only be as fresh as Austin's data, and the feed already ran a day late once, on October 21. And cancellations running hotter than assumed - which is why trial customers get a check-in on day 7 and, on day 14, a recap of exactly what their trial delivered before they are asked to pay.

A prediction, written down where it cannot be quietly edited later: the middle path - 34 or more paying subscribers by June 30, 2027. Stated confidence: 35%. In plain words, a real chance - not a likelihood. November 12 is the first evidence either way.

Bottom line: slower is not failure and faster is not likely; the middle path is the one we plan around.

What could all the businesses do in a year?

In one sentence: pages 14-15 projected Permit Intel alone; this page adds the other businesses we are now running - as scenarios with their assumptions printed. Read the whole page as what could happen if the checkpoints go our way - not a promise.

Every path starts from the same place: $0 of revenue and 0 customers today. The Permit Intel figures are the three paths from pages 14-15, unchanged: Austin first, and a second city joins only in the paths where the November 12 verdict passes. Month 1 of every path starts at $0 - that is where we actually are. The other businesses are younger and smaller, so their assumptions are printed next to them - and Lode Star, which has not yet had a single customer conversation, is counted at $0 in two of the three paths.

Business (assumption)	Slower year	Middle year	Faster year
Permit Intel - $149/mo subscriptions (pages 14-15)	$20,700	$55,300	$120,700
Contra - one-off services at $49-$195 (about 1 / 2 / 4 jobs a month)	$1,200	$3,000	$7,200
Upwork - Eva-run work at $35 an hour, before Upwork's 10% fee (about 2 / 5 / 10 billable hours a week, for most of the year; priced at today's $35 all year - any rate rise as reviews accumulate is not counted here)	$2,800	$7,700	$16,100
F5 - $299 claim-verification offers (0 / 2 / 6 takers all year)	$0	$600	$1,800
Lode Star - language services (translation, transcription, captioning), still at watch grade: real government demand evidence, but zero packets sent and zero customer conversations so far - no evidence base for these numbers yet	$0	$0	$1,500
Monthly income in month 12, all businesses together (approximate: subscriptions at $149, plus that month's Contra jobs and Upwork hours)	about $3,950	about $11,100	about $25,500
Total collected across the year, if that path happens	$24,700	$66,600	$147,300

How to read this honestly. Permit Intel does the heavy lifting in every path - the other businesses are options we are testing cheaply, not pillars the plan leans on. The Lode Star line in the fastest path is the single most speculative number in this report: that business has produced no packets and no customer conversations yet, and it appears only to show what even a small start would add. Every checkpoint in this report still applies - if November 12 kills Permit Intel, its line comes out of all three paths and the totals fall with it. And Glen's $700-a-week living draw is not in these numbers; it comes out of the fund itself, and it is counted where it belongs, in the ask (page 4).

Measured against the $5,000-a-month goal: the middle path passes it during the year (Permit Intel alone crosses it in June 2027 on page 15, and the other businesses only add), and the faster path passes it sooner (March 2027). The slower path does not reach it inside the twelve months - it ends the year at about $3,950 a month, and we would rather show that than round it up. Costs against all of this: about $240 a month today, rising in stages but staying in the low hundreds (next page) - so even the slower path's month-12 income covers the running costs many times over, before Glen's time and taxes.

Bottom line: several small businesses together could reach the goal sooner than one alone - possibilities with the math shown, never promises.

What does it cost to run - and what will you need later?

In one sentence: the whole operation runs today for about $240 a month, and every cost of growing is a planned stage item - nothing on this page is being bought tonight.

What it costs today (October 28, 2026)

Item	Per month
Muse subscription - the service Eva runs on	$100
Lemlist email platform - the held wave's sending tool; its renewal is a decision we make when the wave launches	$69
Rented work server in Germany - one-month trial; verdict November 23	€29.91 (about $33)
Upwork membership, plus proposal credits as used	$19.99
Domains and mailboxes, across several small items (one mailbox alone is $36 a year)	about $15
Website hosting and payment processing - the payment processor only takes a cut when a sale happens, and there are no sales yet	$0
Total running cost today - approximate	about $237

In words: about $240 a month keeps everything running today - less than the $298 that just two subscribers would bring in.

How the cost grows - a ladder, not a leap

• Stage 1 (first paying customers): add backup internet (below) - the running cost rises to roughly $250 a month.
• Stage 2 (marketing running): measured marketing spend comes from the Stage 2 capital, only where returns can be measured; a larger Germany server, if the work earns it, is a step up from €29.91 a month, priced when chosen. Running costs stay in the low hundreds a month.
• Stage 3 (several businesses earning): a one-time computer purchase (below) replaces rented stopgaps. Even grown, the monthly cost is measured in hundreds - against a $5,000-a-month goal.

What we will need as Eva grows

This is the amazing-things list - and every item on it is here because Eva has genuinely run into the wall it removes. Nothing here is a day-one purchase; each item unlocks at its stage, as customers appear.

• A dedicated, always-on computer for Eva - the top of the list. Today Eva's home base is a borrowed slice of a shared system with very little working memory, her heavy workshop work rents space on a server in Germany, and for anything on a real desktop she borrows Glen's own laptop. Her own modest machine (a small mini-PC, always on, hers alone) means she works around the clock without borrowing Glen's computer. The other named path is a larger rented computer in Germany, on the rented route we already use - rented forever, but nothing to break at home. One of the two gets picked at its stage; roughly $500 to $1,000 one-time for the home machine, approximate. The honest trade-off: the home machine is ours and fast to reach; the Germany machine rents forever and costs more per month as it grows, but it never breaks down in our house - and it is already where Eva's workshop lives.
• Backup internet: Starlink. This very evening, home-internet dropouts stopped work cold for stretches - that is the exhibit. Starlink's standby failover is $10 a month (checked October 28, figures approximate; the dish kit is $349 to buy, and full-speed plans run $55 to $130 a month if we ever need more). A planned stage item - no purchase is approved yet.
• A battery backup (UPS), so a power blip does not take everything down - roughly $100 to $200 one-time, approximate.
• Eva's phone presence - a phone of her own, with its own line (device plus line costs, approximate). The phone chapter is built but parked: today Eva borrows Glen's phone by screenshot. A dedicated device unlocks when a real recurring phone job exists - and not before.
• The tools subscriptions, kept current: the AI service that powers Eva ($100 a month today), the email outreach tool, domains and mailboxes, the small rented work server (about $33 a month), and the Upwork membership and proposal credits. This is the running cost on the table above - the business fund keeps it funded; Glen's living draw comes from the same fund, openly (page 4).
• Business formalization, when revenue justifies it: proper business registration, bookkeeping and accounting help, and contract review - the Stage 1 items already named in the $30,000 plan.
• Measured marketing (Stage 2 of the plan): the advertising and outreach budget that is only spent after the earlier stages produce paying customers, and only where returns can be measured.
• A reserve for breakage and replacement (Stage 3): laptops and routers die. The reserve means a dead machine is an errand, not a crisis.

Each item unlocks at its stage, the same rule as the capital itself: customers first, equipment after.

Bottom line: everything runs today for about $240 a month; everything we will need later is listed, priced, and unlocked only by the business's own stages.

Why not the cleaning business?

In one sentence: Growth Cleaning is a real program, but its hardest parts - hiring cleaners, cash timing, and a price you only learn on a sales call - were the wrong hard parts for our family.

Fairness first. Growth Cleaning is a real coaching program, run by Michael Haeri, who operates his own cleaning company. It teaches a "remote cleaning business": you hire local cleaners, sell over the phone, and manage everything from a laptop - you never clean yourself. It has coached more than 400 clients, its reviews are mostly positive, and some of its students clearly do well. I am not calling it a scam. I am saying its challenges were the wrong ones for our family. Here they are, in the open:

• The price is only revealed on a sales call. The program does not publish its enrollment fee - its own site says you "will get the real number" on the call, and its reviewers describe the course as "not cheap." A large share of the $30,000 would have been committed up front, before the first customer existed.
• The headline numbers are revenue, not profit - by the program's own math. Their materials show one cleaner bringing in about $4,000 a month, of which the owner keeps roughly half - before software, insurance, advertising, payment fees, taxes, and the owner's own hours. Their "$10 million in client sales" figure carries their own fine print: "Individual results. Not typical."
• Cleaners are the business, and they are the hardest part. The program's own sales page lists the industry's problems: "no-shows, theft, insurance headaches, high turnover." Experienced operators plan on replacing a third to half of their cleaners every year. Hiring is never finished - and when a cleaner quits mid-week, the clients still expect clean homes, and the owner is the backup plan.
• You answer for work you never see. The owner never enters the homes, so quality is managed through photos, checklists, and complaints. Operators say quality slips within weeks without constant checking - and every bad clean carries the business's name, not the cleaner's.
• Real liability, from day one. Cleaners work inside strangers' homes: liability insurance (roughly $500 to $1,500 a year), bonding, background checks - and a single incident, a broken television or a flooded bathroom, can run past $10,000.
• The cash flows the wrong way while you grow. Cleaners expect to be paid weekly, so growing means paying people ahead of the revenue. Experienced operators recommend a reserve or a credit line for exactly this gap - more family money at risk, for longer.
• The easy hours come later - if they come. The program's own FAQ asks for real hours through the first 90 days, days spent on the phone with customers, and warns of "a few uncomfortable months at the start," with 6 to 12 months to reach full scale. The five-hour weeks in their stories belong to mature businesses with staff and systems - the destination, not the starting point. With three young children, my free hours are evenings. The first 90 days of this business are daytime phone calls and people problems.
• I have already run this race. I owned a carpet cleaning business and sold it in January. I know firsthand what cleaner reliability, scheduling, and customer complaints do to a family's week. The remote version changes where the owner sits. It does not change what breaks.

The contrast, in one place

• Growth Cleaning: the whole $30,000 deployed up front, from the first week; income capped by Glen's physical hours; a local service business, worked in person, every day.
• This operation: the same $30,000 deployed in stages, behind pre-set checkpoints; income not capped by one man's hours; and if the kill numbers hit, the undeployed stages simply remain unspent.

None of this makes Growth Cleaning a bad program for somebody else. It makes it a business whose hardest parts land exactly where our family is thinnest - time, and cash committed up front with no cheap way to test first. This operation asks for the same $30,000 Growth Cleaning would have taken - but staged, so it is spent only after customers prove each step, and the first proof lands by November 12 for under $200, before any of the capital is spent.

Bottom line: Growth Cleaning would have spent the whole $30,000 in the first weeks on a business capped by Glen's hours; this asks for the same backing, staged behind proof.

What could go wrong, and what we do about it

In one sentence: the risks are real, each one has a plain answer, and the biggest one - whether anyone wants this - is exactly what the test is for.

Risk	How we handle it
Demand is unproven until November 12	That is exactly what the capped test is for: success and kill lines set in advance, total exposure under $200.
New competitors keep entering	We compete on the adjacent-trade niche, price, and service, and on being first in each city - not on pretending the data is secret.
Austin could stop publishing its free feed	Many U.S. cities publish the same data, and the system is built to switch cities. No single feed is a single point of failure.
Email marketing rules	Outreach follows the federal CAN-SPAM rules, opt-outs are honored, and we never sell homeowner data.
It depends on Glen and Eva	If either of us is unavailable, the business pauses - it does not collapse. The assets (subscriber list, data pipeline, brand) remain in place.

Questions we expect

Why would contractors pay for data that is free?
Because "free" costs them hours. Digging through the city's portal every week is real work. We sell back their mornings - the same records, already sorted by their trade and their ZIP codes, ranked, in their inbox at 7 a.m.

Is this legal?
Yes. The data is published public record, made available by the city for anyone to use. Our outreach follows federal CAN-SPAM rules, and we never sell homeowner contact information.

Who runs it day to day?
Eva operates it - the data, the briefs, the outreach, the follow-up. Glen owns it and makes the money decisions: pricing, spending, and growth.

What happens if it fails?
The test stops on November 12 having spent under $200 on the test itself. Spending from the capital fund stops, the unspent balance stays in the business account, and Glen reports the outcome to the family plainly. We will have bought a clear answer, cheaply.

How will the family know how the money is being used?
The capital sits in the business account, spent only against the itemized plan on page 4. Glen reports results as the milestones land - the November 12 verdict first, then subscriber and revenue numbers as they happen.

What exactly is Eva - and what can she not do?
Eva is the AI operations partner who runs the day-to-day work under Glen's ownership. She cannot spend a dollar without Glen's own approval, cannot sign anything, and cannot announce a number the payment records do not confirm - the checking system verifies her work every night. She is also still in training on Glen's laptop: she has not passed her written driving exam yet, and some surfaces are hard-blocked to her entirely.

Do we get it back?
It is a gift, so no - and Glen will not dress that up. What he promises instead: the staging, so most of it never moves unless customers exist first; an update after every checkpoint, with the balance in it; and that the undeployed stages are never quietly spent - the one household use of this fund is Glen's $700-a-week living draw, stated openly and counted in every update. If a kill verdict ends the operation, whatever remains unspent stays visible in those updates, and Glen will say plainly what is left and talk it through with you.

Why does an online business need $30,000 if it costs about $240 a month to run?
It does not need it to run - it runs today for about $240 a month. Part of the $30,000 is Glen's living draw - $700 a week while the businesses get going, about ten months of runway from the fund alone. The rest is mostly reserve and staged marketing that only deploys behind paying customers. It buys speed and staying power for something already built.

Who decides to stop?
The numbers written down in advance decide - the kill lines on each business (November 12 for Permit Intel, November 25 for Contra, 20 proposals with no real interest for Upwork) - not moods, not hope, and not one bad week.

Bottom line: the risks are real, written down, and each has an answer - and the kill numbers, not hope, decide when to stop.

One last thing - and an invitation to talk

In one sentence: we are asking for $30,000, released in stages that only paying customers unlock, spent only on the plan in this report, with results reported to you as they happen.

The same $30,000 you were ready to provide for Growth Cleaning - for a family of businesses that is already built, already testing, and costs about $240 a month to run (updated October 28).

To be clear about what you would be funding:

• One amount, in protected stages: $30,000. It goes into the business account and is spent only against the staged plan on page 4: about $5,000 for post-verdict expansion and legal setup, about $10,000 for marketing tests with measurable returns, and about $15,000 held as operating runway and reserve. Each stage unlocks only when paying customers from the stage before it exist - and if November 12 hits the kill criteria, nothing deploys at all.
• It is a gift, plainly. No loan, no repayment schedule, no equity, no ownership stake. The stages and stop lines are Glen's own discipline, offered freely - not terms the family set.
• Part of it is Glen's living draw. $700 a week comes out of this same fund for the family while the businesses ramp - stated openly, and counted in every update. The rest exists to grow businesses that are already proving themselves.
• It buys acceleration and time. The businesses exist without it. The capital buys more cities sooner, proper legal foundations, real marketing tests, and the breathing room to build subscriptions without income pressure.
• Accountability is simple and direct. Glen reports results to the family as the milestones land - starting with the November 12 test verdict, shared either way, good or bad - and then subscriber and revenue numbers as they happen.
• The test's honesty stands regardless. The test itself cost under $200, its pass and fail lines are written in this report, and if it fails, spending stops and we discuss the unspent balance openly.

One thing should be said plainly: you were ready to back me on the cleaning business, and stopping that was my call - you were backing your son, and that has never been lost on me. This report is me asking you to place that same backing where almost none of it moves until customers exist first.

Andrea, Mom and Dad - thank you for asking what I am working on, and for reading this far. I would rather show you a small, honest test with its failure lines written down than a big, confident pitch. That is the whole spirit of this business: prove it small, prove it cheap, then grow what the evidence supports.

Whatever you decide about the capital, the test reports back on November 12 either way - and you will hear the result from me first, good or bad.

When you have had time with this, let's sit down and talk it through - a real conversation, not another document. Mom, you asked for exactly that back in October, and you were right to. There is no deadline on your decision and no pressure from me; the test reports on November 12 either way, and I would rather answer every question out loud than send one more page.

- Glen

Permit Intel · trypermitintel.com · $149/month, month to month · Austin, Texas market · Report prepared October 21, 2026. All Permit Intel figures in this report are as of that date: $0 revenue, 0 customers, $199.16 spent or committed, 294 verified Austin prospects. Competitor prices and company history are drawn from published company information as researched in October 2026. Projections are illustrative scenarios at the stated price, not forecasts or guarantees. Updated October 28, 2026: the update on page 3 and the reframed $30,000 staged ask (pages 4 and 20) reflect the whole operation as of that date; the $0 revenue and 0 customers scoreboard is unchanged.

Bottom line: thank you for backing your son. The invitation stands: let's talk - no deadline, and no pressure.