On Top Home Services LLC · Salem, Oregon
Four systems into one
On Top Home Services is my exterior cleaning company. For years it ran on WordPress, Smartsheet, Google Drive and DocuSign — four subscriptions, four renewal dates, and me acting as the integration layer between all of them. Here is what that cost, measured from my own accounts, and what replaced it.
The tools were fine. The seams were not.
Nothing on this list is bad software. The cost was never any one of them — it was the gaps between them, which a person was quietly filling by hand.
- 01WordPressPublic site and contentA theme plus plugins; every change routed through someone else.
- 02SmartsheetJobs, scheduling, accounting, mileageA grid pretending to be an operations system, re-keyed from quotes.
- 03Google DriveQuotes, estimates, invoicesThe real system of record — loose documents, searchable by nobody but their author.
- 04DocuSignSignaturesPer-envelope pricing on documents assembled by hand elsewhere.
The tell was always the same question: “what did we invoice last month, by service?” It could not be answered without a spreadsheet session, because no single tool held both the job and the money. That is not a reporting problem. It is a structural one.
Customer names, phone numbers, filenames and document contents are redacted in the three screenshots above. They are live production tools, and the redactions are left visible rather than cropped out — a cropped screenshot asks you to take it on trust that nothing was hidden.
Consolidation, drawn
The argument in one picture: four tools each holding part of a customer, replaced by one system where every module reads the same record.
What one system actually bought
01
One customer record
Captured at enquiry and carried through to the reconciled payment. Nothing is typed twice.
02
Quoting inside the system
Drafted from job type and service history rather than from a blank document.
03
Signature in the flow
Approval happens on a phone, against the quote that generated it, with no envelope fee.
04
Scheduling that knows the job
Crew and date assigned against the same record, visible to everyone who needs it.
05
Invoicing on completion
Raised from what was actually done, not reconstructed from three places a week later.
06
Content the owner controls
Service pages, city pages and media edited in-house, without an agency ticket.
From a WordPress and multi software headache to a simpler solution
My rebuild carried the local search structure now running across 17 Willamette Valley cities — a real page per service, a real page per city, and dense internal linking between them. That structure is the reason a home services company competes with national lead-generation brands in towns the size of Willamina, and it is the same structure this site runs on. But now I have auto-quoting available, easy reporting on everything, and my customers can easily auto quote and see their job details.
It took me months to build in WordPress and supporting apps what this system can provide in a week. That is the difference between assembling a business out of plugins and owning the thing outright.
What running my business actually costed before the upgrade
From my own accounts, not a survey. If you run a contracting business these figures will look familiar — the point is that most owners have never seen them written down.
Average ticket
$726
The average across 689 completed and paid jobs — what the business actually banks per job.
689
jobs, completed and paid
Every job that finished and collected money, 2020 through 2026. Not a rounded-down floor and not a record count — the same 689 jobs every other figure on this page is drawn from.
2 hr
per job, administrative and marketing
Quoting, scheduling and paperwork on a typical job. Driving is measured separately and taken out, and so is the build time I logged against jobs while writing the software — what is left is desk work. My own figure rather than a raw timesheet total, which reads higher for exactly that reason.
$82,036
spent for 3.0 yrs (FTE) to be found
Marketing, the website and digital advertising, 2020 through 2025 — one heading in my own accounts, not a bucket assembled to make a point. It averages $27,345 for every full-time year of my own effort it took to earn.
Set against what it produced — around $470,000 of work across 2020 through 2025, the same years that spend covers — that is roughly 17% of gross revenue spent on being found, or about $5.70 back for every dollar in — a 5.7× return on marketing spend (ROMS). For a growth-focused trade business that is a healthy and ordinary ratio, provided the gross margin supports it. And it did not buy advertising alone — the same money built the operating system the whole business now runs on, end to end.
Two levers move those hours, and they are different jobs. Automation takes out the re-entry — quoting, scheduling and invoicing off one record. Scheduling takes out the driving, by offering the customer the dates a crew is already routed near them rather than the next open slot. Most quoting tools do the first and none of them do the second.
Same trade, same territory, better work
Measured, not forecast. One number in six years of records moved enough to be worth putting on a page — and it is not the one I expected.
$609
2020 – 2022
Average job before any of this existed. WordPress, a spreadsheet, and quoting from a blank document at the kitchen table. Before the top-shelf engineered washing machine.
$759
2023 – 2024
The rebuild years. The system going in, the search structure going up, and the work starting to arrive differently. After the top-shelf engineered washing machine — and moving into property improvement.
$855
2025 – 2026
Same trade, same territory, same truck. An average job worth 40% more than the one I was doing five years earlier. My chargeable value increased because of consistent quality of service increase as I built continued trust in the market.
These numbers show I know how to grow a business. I want to help get you there quicker so you don’t have to go through all the growing pains I had to — you can have in one year the web presence that cost me three full-time (FTE) years to build. Forty per cent increase in value, across 689 jobs. I know how to read a market and help the numbers work for you. In my first three years, paid lead generators brought me 37% of my work. Today it is 15%, and my own website and returning customers bring three quarters of it. A trustable web presence brings more jobs, and a solid operating system helps you manage what matters.
Six years of analyzing the market and building my own software, during downtime, is a benefit you can prosper from.
Where the hours go in the next 3 years (projected)
Everything above is measured. This is not — it is a forecast, resting on one assumption I am willing to state plainly and be held to: that quoting a job automatically, and having tools to administrate the life cycle of the record, halves the office time each job costs me.
1 hr
per job, projected
Half of the 2 hrs of office work a job costs me today. Quoting, scheduling and invoicing stop being three acts of typing against one customer and become one record moving through a system. This is the assumption doing all the work on this page, and it is an assumption.
40 Days
back every year
Not saved money — returned time, which can go to field work, to a second crew, strategizing, or to off time.
120 Days
across the three years
Four months of working days, handed back to a business that never had them to spare.
I am not promising the advertising bill drops. What changes is what it buys. A page that ranks keeps ranking while I am up on a roof, and a search structure built once does not need re-buying every month the way ads do — so the same twenty-seven thousand a full-time year starts buying capacity instead of attention. When you don’t have to spend hours on the phone with GoDaddy, or months to do what now takes weeks, you know things are streamlining.
If the office time does not halve, I will say so right here. That is the deal — this section gets rewritten against real numbers, never quietly left up. Everything above it is already measured. What I would like to do next is build the same thing for somebody else, and get them there in a good deal less than six years.
You already know most of this
None of it will surprise you. The point is that you can’t manage what you don’t measure.
You are the integration layer
Every tool you pay for works. What costs you is being the only thing connecting them — and you cannot take a week off from that job.
The expensive hours are invisible
Nobody invoices for quoting, chasing, scheduling and driving. It never lands on a profit and loss statement, so it never gets managed.
Advertising is many times an afterthought
It gets whatever attention is left once the work is done and the paperwork is filed — which is backwards, because it is the part of the business that decides whether there is any work next month.
What you see is what you get
A trusted partner with skin in the game, who wants to help save you from some unnecessary blood, sweat and tears.
The screenshots above are real production screens, and the customer data in them is destroyed rather than hidden — pixelated at the source, because a blurred name can be sharpened back and a cropped one only asks you to trust the crop. What went is listed plainly: every filename in the Drive list, the name and phone columns in the Smartsheet grid, and the quote document in DocuSign. The column headers were left, because they show what the sheet held without showing whose it was. The operations dashboard needed no redaction; it carries no customer data at all.
On Top Home Services is not an arm’s-length client: it is my company, which is exactly why the system could be rebuilt from the ground up rather than negotiated in increments. Treat it as a worked example, not an independent reference — and ask for one of those separately.
Start with it already built
For exterior cleaning companies — pressure washing, roofs, gutters, windows, lights. No minimum size, because the argument works from either end: if you are already running, this pays for itself as you work, and if you are just starting, you make the investment and start on a good foundation.
$25,000
the build, yours outright
Operating system, website, search structure, Google Business Profile and social set up and verified, and the reminders, review requests and channel tracking wired in. Delivered in the first quarter and handed over. It costs this rather than what custom software normally costs because it is not being invented — it already runs an exterior cleaning company.
$34,000
the first year, looked after
The build, plus $1,000 a month for months four to twelve — expansion into the towns that turn out to convert, monthly checks, a page a quarter, and the channel report throughout. Search compounds slowly, and the first year is when leaving it alone costs the most.
$250/hr
After that, if you want it
Believe me, I get a lot done in an hour. Also, by then you know how to do everything and don't really need me. Nothing renews automatically. No retainer, no notice period, no platform you are locked into — you own the code, the domain, the data and the hosting account from the start.
For scale — what the same three things cost bought separately from a competitor, taking the bottom of every published range rather than the top:
- Custom field-service softwarescheduling, quoting, invoicing, one customer recordfrom $25,000
- A custom websiteagency build, not a templatefrom $5,000
- Local search worktwelve months at contractor ratesfrom $18,000
Forty-eight thousand dollars, at the floor of every range — and at the end of it the software is either rented or bespoke rather than yours. This is $25,000 for all three, owned outright, or $34,000 with the year of upkeep. Ranges from 2026 industry pricing surveys; I have quoted nobody by name and nobody’s numbers but my own.
Alongside either one I would put $1,000 a month into the applicable lead generator or ad platform while the search structure matures — but that is your spend, billed by them, and it carries no markup from me. It is a bridge on purpose: around three months in it is yours to run, or to stop. If you would rather I stood the accounts up, that is $300 per account — or set them up yourself, they are in your name either way.
The system runs on a few outside services — hosting, database, email and the AI tooling behind the quoting. Those accounts are opened in your name and billed to your card from the day the build starts, not routed through me and not marked up. Budget somewhere around $50 to $150 a month across all of them at the volume most exterior cleaning companies run at. I set them up with you, which is what makes owning the thing outright mean something.
That first year is the handover. What it buys is the thing that cost me three full-time (FTE) years and $82,036 to work out — built once, owned outright, and running before you have paid for the mistakes I already made — and you pay half of that at most.
What the year actually looks like. Most of the work is in the first quarter, by design — the system and the search structure get built once, and the rest of the year is keeping them running.
- Months 1–3The buildOperating system installed and your records moved in. Website and search structure live — a page per service, a page per city you actually serve. Google Business Profile and social verified. Seasonal reminders, review requests and channel tracking wired in. Ad bridge switched on.
- Months 4–6Tune, then hand overThe ads become yours once they are steady, usually around month three. City coverage extends toward the towns the data says convert rather than the ones we guessed at the start. Anything the first quarter proved wrong gets fixed.
- Months 7–12Keeping the lights on and helping you growMonthly: indexing and rankings checked, profile kept current, anything broken repaired. Quarterly: a new service or city page where there is demand to justify one. The channel report runs throughout, so you can watch paid work shrink and owned work grow.
Next step
How many systems is your business holding together by hand?
The audit that starts this work is fixed-scope and ten business days. It ends with a ranked register and a roadmap naming the first build.