Do nothing
Your calculator number, every year, forever
The cheapest one on paper. It is also the only one where the bill never stops.
For small businesses, nonprofits and schools
Where is your business losing money?
In two places. Your team does work a computer could do. And your prices, your costs and your customers were never checked. I find both. Then I fix them.
564 hours a year given back on one job · $25M+ found · 5 years doing this
What you get, before you pay anything
- 1A 30-minute callOften the same week
- 2A written plan, with your numbers in itWithin 3 business days
- 3A build, only if the plan is worth doingPriced per job. If I am late, you do not pay the rest.
The first two are free. You keep the plan even if you never hire me.
Where your week goes
Most businesses run on four or five apps. None of them talk to each other. So a person has to move the numbers by hand. That person costs you more than the apps do.
Sales
Invoices
Inventory
Payroll
A person
retyping it all by hand
A report, two days late
Four apps and one person in the middle. Every number gets typed twice. Mistakes turn up weeks later. And it all stops when that person is out.
Sales
Invoices
Inventory
Payroll
One pipeline
nobody watching it
A dashboard that is always current
The report, in your inbox on the 1st
You, deciding
Now the apps talk to each other. Nobody types anything twice. The numbers match. And the person in the middle reads the answer instead of building it.
What changes
Same business, same people, same software. Different plumbing.
Someone types numbers from one app into another every Monday.
The apps talk to each other. Nobody types it twice.
The monthly report takes two days. By the time it lands, the month is over.
It builds itself overnight and lands in your inbox on the 1st.
Finance, operations and the board deck each have a different number.
One number, one source, and you can show where it came from.
Invoices and follow-ups go out when somebody remembers.
They go out on time, every time, without anyone remembering.
You do not know which jobs or customers make money.
You know by job, by customer and by product.
A new question means a two-week job to get an answer.
You ask, and the answer is already there.
What is it costing you?
Five common leaks, run on your numbers instead of mine. Pick the one that sounds like your week.
The task nobody questions
Someone spends hours each week on work a computer could do. It never shows up on a bill, so nobody ever chose to spend it.
$
The arithmetic
- True cost per hour (wage x 1.3)
- $45.50
- Hours lost per week
- 24 hrs
- Cost per week
- $1,092
a year, on work nobody chose to do
Pay times 1.3 covers taxes, benefits and time off. That is a low estimate.
The appointments that never arrive
A no-show is a slot you paid for and a sale you counted on. Reminders will not save them all. They do not need to.
%
$
The arithmetic
- Appointments per month
- 20
- Missed per month
- 2
- Missed per year
- 24
a year, walking out before it walks in
Based on 20 work days a month. Save half of them and this pays back faster than anything else here.
The errors you find too late
A wrong number typed on Monday gets caught at month end. Or by the customer. Someone has to undo it either way, and the customer one costs more.
%
$
The arithmetic
- Errors a month
- 12
- Errors a year
- 144
- Cost to unpick each
- $120
a year, fixing things that were entered wrong
Count the whole cost: finding it, fixing it, and making the customer happy again. Most people guess low here.
The customers you did not have to lose
You will lose some customers no matter what. The rest go quiet and nobody notices, because nothing is watching.
%
$
The arithmetic
- Customers lost a year
- 24
- Value of each
- $5,000
- Still on the books
- 176
a year walking out, some of it avoidable
You cannot save all of it. But you can see how big it is, and spot who is drifting while there is still time to call.
The increase you have not made
Raising a price is easy. Knowing who will pay it and who will walk is the hard part. That is why most businesses leave it alone for years.
$
%
The arithmetic
- Revenue
- $2,000,000
- Increase of 3%
- $60,000
- Extra cost to deliver it
- $0
straight to profit, because your costs do not move
This assumes nobody leaves. Checking that is the job: who can take a rise, and who cannot.
You can price these in ten seconds. The ones hiding in your data are often bigger.
Book a 30-minute callSystems scale. People do not.
Twice the orders means twice the hand work. The only way to do twice as much by hand is more hands. A system does not care if it runs 40 orders or 400.
Manual processAutomated system
Hours of work per month04080120160
145 hrs38 hrs107 hrs a month
you would be hiring for
you would be hiring for
Today2x3x4x5x
Business volume across the bottom. This shows the shape, not real data. Your line may be steeper or flatter. The gap still opens.
- The manual line is a hiring plan
- Each step up that line is another wage, or the same people staying late. One costs money. The other costs you the people.
- The flat line is the whole point
- Build it once and growth stops adding work. The gap between the lines is what you keep.
- Build it before you need it
- Fix it while it is small enough to follow. At five times the size it is the same job, but it costs more and you need it sooner.
What I build
Two halves of one job. Most work starts with automation, because the hours it saves pay for the rest.
Automation
The waste you can seeI watch how the work really gets done, not how the handbook says it should. Then I rebuild the boring parts so they run on their own.
- Systems that do not talk to each other, connected
- Reports that build and send themselves
- Forms and intake that write straight into your records
- Invoices, reminders and follow-ups that go out on time
- Data cleanup that runs all year, not once a year in a panic
Analytics
The waste you cannotI pull your apps into one clean set of numbers. Then I answer your question in writing and tell you what to do. Charts if you want them.
- Which customers and products make money. Most owners guess wrong.
- What is holding your growth back, measured instead of guessed
- Which supplier, material or customer could shut you down next month
- Which prices can move without losing the customer
- Checks that catch bad data before your customers find it
- Written answers for people who need a decision, not another dashboard
Margin by product
| Product | Revenue | Margin |
|---|---|---|
| Product A | $412k | 31% |
| Product B | $288k | 24% |
| Product C | $196k | 9% |
| Product D | $174k | -4% |
Product D sells well and loses money on every unit. This is the kind of thing that stays invisible until somebody measures margin per line rather than in total.
Hours of manual work, by month
Manual work grows with the business. Headcount is the usual answer, and it is the expensive one.
How it works
You get the numbers and the plan before you pay for anything. If you like the plan and want to do it yourself, take it and go.
- 01Free
A 30-minute call
Often the same week
You tell me where the time goes and what you cannot see.
No slides. No pitch. I ask four questions. You find out if there is enough here to be worth our time.
- Where does the week go?
- What can you not see right now?
- What apps does your data live in?
- What is it worth if it goes away?
- Thirty minutes, one call, no follow-up sequence
- If it is not a fit, you hear that on the call
- 02Free
Your audit and action plan
Within 3 business days
The numbers, written up, and yours to keep.
I look at what you run on. Then I come back with where the money goes, what to fix first, and what each fix is worth. In writing, with the math shown.
- The full math in writing, not a summary
- Where the hours and the profit go
- A ranked list: what to fix first, and what each fix returns
- Yours whether or not you hire me. No conditions.
- 03Quoted per company
The build
Timeline set per project
Only if the numbers make sense to you.
I price it and plan it for your business. A two-person office and a thirty-person shop are not the same job. The price and the date are set in writing before I start. Neither one moves.
- Fixed scope and fixed price, agreed before anything starts
- Delivered by the date in your quote, or you do not owe the balance
- Everything runs in your accounts, not mine
- Notes your team can follow
The first two steps cost you half an hour and give you a priced plan. That is the whole risk.
Book a 30-minute callThe guarantee
Late means you do not owe the balance.
Every price comes with a date, and we agree it before I start. The clock starts the day you give me access, not the day you sign. If it is not running and doing what we agreed by that date, you do not pay the rest and your deposit comes back.
The price does not move
We agree the job in writing first. No hourly billing. No surprise charges at the end.
You own all of it
I build it inside your accounts. If we never speak again it keeps running, and your team can look after it.
I will tell you when it is not worth it
If it will not pay for itself, you hear that on the call. I will not sell you a job you do not need.
Results
Nine problems I have already solved. Most of my work was inside big factories: 20 plants and supply chains with thousands of moving parts. The problem does not change when the company gets smaller. A 12-person shop typing invoices into two apps has the same problem as a giant firm typing into two big systems. Same fix. Smaller and cheaper.
Startup, small business, school, nonprofit, big factory. The trade changes. The apps change. The budget changes. The problem under it stays the same.
- a year handed back, one project
- 564 hrsa year handed back, one project
- in business value identified
- $25M+in business value identified
- Fortune 500 to small business
- 5 yrsFortune 500 to small business
Open whichever one sounds like your business.
Small businesses
The two that look most like you. Neither came with a big budget or a data team.
NFL brandwhat the startup grew intoWhat does this look like for a business my size?
Clothing, ecommerce then retail
What was happening
A small clothing startup selling online. Decisions got made the way they do in every early business: on instinct, on what sold last week, and on whatever the platform dashboard happened to show that day. Marketing money went out with no reliable way to tell which of it came back.
What I built
The pipeline first, so the numbers lived in one clean place instead of four. Then models on top of it, rebuilt and refined over years as the business changed shape: ecommerce first, then retail alongside it with completely different economics. Profitability, marketing attribution, inventory and demand, each one honed until it held up.
What changed
Profit became something they could predict rather than discover at the end of the month, and the marketing numbers became consistent enough to act on. Over that stretch the business grew from a startup into a brand supported by an NFL team.
What it means for you: This one is closest to your business. Same tools as the big jobs, at the size you run at, built up over years instead of dropped in at once.
Cattlegenetics, breeding, forecastingDoes any of this work outside an office?
Cattle genetics and breeding
What was happening
Ranches breeding Wagyu and other premium cattle make genetic decisions that pay off years later. The feedback loop is long, the variables are many, and most of it runs on experience and instinct because that is what has always been available.
What I built
Analytics across breed and genetic data at top ranches, then forecasting for the breeding decisions themselves: which pairings, which lines, and what the outcome is likely to be years out.
What changed
The work went deep enough into what drives quality that it ended with me judging a national steak contest. That was never the goal. It is just where the data led.
What it means for you: If your business runs on one experienced person making the call, that call can be measured, tested and repeated. You do not have to work in spreadsheets to have a data problem.
Schools and nonprofits
Different words, same shape. Money comes in, money goes out, and something in the middle is not paying for itself.
QuickBooksread per program, not in totalWhat if we are not a business at all?
Private school finances
What was happening
A private school with everything already sitting in QuickBooks and no way to read it. Tuition, financial aid, program costs, fundraising and facilities all posting into the same set of accounts. That is fine for filing and useless for deciding. Nobody could say which programs paid for themselves, which quietly ran at a loss, or what the discounting added up to across a year.
What I built
I took their QuickBooks data and rebuilt it into something answerable: revenue and cost lined up by program and by category, the way a business looks at margin by product, with the year-over-year movement next to it.
What changed
The money can now be read one activity at a time. What a sport costs against what it brings in. Whether the cafeteria makes money or loses it. Which lines drain cash and which carry the rest. I am not claiming a savings number here, and that is the point. What they got was knowing exactly where to look next, which nobody had before.
What it means for you: Your accounting software already holds the answer. It is set up for your accountant and the IRS, not for you. That is why the number you want is never the one on the screen. Same job whether the line is a sports team or a product range.
Global manufacturers
Bigger tools, same problems. This is where I learned the methods before pointing them at smaller businesses.
50 → 3 hrsthe monthly cycleWhy does the first week of every month disappear into a report?
Monthly planning cycle
What was happening
The monthly demand plan ran on spreadsheets. Analysts pulled exports from several systems, reconciled them by hand, and rebuilt the same workbook every month. It took about 50 hours, more than a full working week, and they spent longer assembling the numbers than reading them. Forecast accuracy at item level was around 9%, so the output was not reliable enough to plan against even after all that effort.
What I built
I replaced the manual preparation with SQL extraction, automated transformations, and a reporting layer that refreshes and distributes itself. Standardized forecast-accuracy views went in alongside it, plus quality checks that flag exceptions instead of waiting for somebody to notice them.
What changed
The cycle went from roughly 50 hours to 3, a 94% cut that hands back about 47 hours every month, or roughly 564 hours a year. Forecast accuracy went from about 9% to over 60%, so the numbers arrived faster and became worth trusting at the same time.
What it means for you: If somebody in your business loses the first week of every month to a board pack or an owner report, this is the same job at a smaller scale. The hours are smaller. The 94% is not.
$25M+in profit nobody could seeWhich products make you money?
Profitability analysis
What was happening
They could see sales by product and by machine. They could not see what each one left behind once the work, the space and the cost were taken off. So calls about what to push, what to buy and what to drop were made on sales alone, which is the number least likely to show you the profit.
What I built
Power BI dashboards combining production, operational and financial data into one model, ranking top and bottom performers by contribution rather than by volume, with drill-down from the whole business to a single product.
What changed
Over two years this work helped find more than $25M in value. The awkward finding is the same one most times: some of the busiest lines make the least money, and nobody could see it because sales looked fine.
What it means for you: Nearly every owner I speak to is sure which customer or product pays them best. Plenty of them are wrong. And the ones who are wrong tend to work hardest on the thing that pays least.
20%between two systems, unnoticedWhy do two people here have two different numbers?
Two systems, two truths
What was happening
Two planning systems reported demand for the same period. One said 2,843,999 units. The other said 3,554,992. Nobody had noticed, and every downstream decision, purchasing, capacity, allocation, was running on whichever number the reader happened to open.
What I built
I traced the discrepancy back to its cause. Planners were making offline adjustments after the consensus forecast locked, so the operational source drifted from the official one every cycle. I changed the process to read the adjusted source where it was the right one, instead of assuming every system held the same figure.
What changed
A 20% error in the number the business planned against, found and corrected. Nobody had reported it as a problem, because from inside either system everything looked fine.
What it means for you: This is the leak you cannot see. It is the same failure as your accounting software and your operations spreadsheet disagreeing about last month, with everyone quoting whichever one they happen to have open.
95%of the retyping, goneWhy does someone spend half their day retyping between two screens?
Cross-system data entry
What was happening
Keeping records matched across two systems meant a person pulling the data, working out the rules by hand, opening a second app, finding each record and typing the changes in. It ran for days. Dull work, but not safe to rush, because one wrong entry spread everywhere.
What I built
A Python automation that prepares the source data, applies the mapping and segmentation rules, drives the remote application, finds the records, fills the fields, and logs or skips anything unusual. It works through the existing interface and existing logins rather than demanding new system access, and it can be stopped and resumed.
What changed
Over 95% of the steps run without a person touching them. The remainder are the exceptions, which is exactly where a human should be looking. A process measured in days became a repeatable workflow.
What it means for you: A barcode scanner instead of typing every code. The person still supervises and still approves. The machine does the typing.
780/yrorders, none touched by handWhy is the same four-step job done by hand on every order?
Billing and reconciliation
What was happening
A recurring billing cycle meant creating each order, assigning the right inventory batch, posting the goods issue and generating the invoice. Four linked steps, every order, every month, tied to month-end close and financial controls, so it had to happen on time whether or not anyone had capacity.
What I built
An automation chaining all four steps together, plus the troubleshooting to make it survive month-end edge cases like quantity and date changes mid-run.
What changed
About 65 orders a month run through it, roughly 780 a year, without the manual transaction work. Users have since identified another 39 a month that could be brought in, which would take it past 1,200 a year.
What it means for you: If any part of your month-end depends on somebody remembering to do the same few things in the same order, this is that, and it is the easiest kind of work to hand to a machine.
4 of 23agreed what a month even wasWhy does nobody agree on what last month was?
Global production reporting
What was happening
Production figures were reported locally by each site and compared centrally. On the first validation only 4 of 23 reported items matched the automated figures. The gaps were not coding errors. Sites disagreed about what counted as production: whether relabeling counted, whether reversals were removed, whether repackaging was included, which posting period defined a week.
What I built
An automated report pulling production orders, material and plant data, capacity and utilization into standard site-level comparisons, with views matching the format the operating teams already used, so every mismatch could be examined at order level instead of argued about in summary.
What changed
One location was reconciled to a documented weekly figure of 489 metric tons. Another was narrowed to a 9 metric ton residual. More than 20 production assets came under one reporting process. The reconciliation was the valuable part, because it surfaced that the business had never agreed on what it was counting.
What it means for you: If two people in your business can produce different numbers for the same month and both defend them, you have this problem. It is almost always definitions rather than arithmetic.
I describe clients instead of naming them. I can show the math behind every number here.
What are you comparing it to?
There are four ways to deal with this. Three of them cost more than you think.
Hire an analyst
$100,000+ a year, loaded
Wage, taxes, benefits, a laptop, your time managing them, and three months before they help. Most businesses under $10M cannot keep one busy.
An agency retainer
A monthly bill with no end date
You rent it instead of owning it. Stop paying and it stops working, because it lives on their computers.
This
One fixed price, once
It runs in your accounts. Your team looks after it. It keeps working whether or not we speak again.
Why this goes wrong, and what I do about it
Most people have been burned by a consultant or a software project. Here are the reasons, in the order they come up.
The last consultant disappeared for three months.
Fixed delivery date, and you stop paying if I miss it.
The price started at one number and finished at another.
Scope and price agreed in writing before day one. No hourly billing.
We got a dashboard nobody opens.
You get written answers with a recommendation. Charts are optional.
It broke, and the only person who understood it had left.
Documentation written for whoever maintains it, not for me.
It lives in their system, so we are stuck with them forever.
Everything is built inside your accounts. You own all of it.
Nobody on our team has time to learn another tool.
The best automation is invisible. Most of this runs with nobody watching it.
We do not know if it is even worth doing.
We do the arithmetic on the call. If it does not pay back, I say so.
Our data is a mess, so we are not ready yet.
Cleaning it up is the job. Nobody who needs this has tidy data.
You are hiring one person, not an agency
I spent five years building reporting inside big factories. Huge supply chains. Twenty plants. The kind of place where being wrong by 20% is a real problem. I spent a lot of that time watching good people lose their mornings typing numbers from one screen to another.
Small businesses have the same problem and none of the budget. There is no data team to ask. There is no one to file a ticket with. And there is no case for hiring one. So the work never gets done, and what it costs you never shows up anywhere you would look.
That is the gap I work in. You get one person who has done this at a size you will never need, doing it at the size you have.
Questions
Including the ones people save until the call.
Why not just hire someone?
A full-time data person costs well over $100,000 once you add benefits, and most businesses under $10M do not have enough of that work to keep one busy. You need the skill, not the extra head. That is the whole reason I exist.
Your experience is all big companies. Does that translate?
The tools are bigger. The problem is the same. A giant firm typing records into two systems and a 12-person shop typing invoices into QuickBooks and a spreadsheet have the same problem, and it gets fixed the same way. What the big jobs taught me is what good looks like: written down, tested, and built so it does not fall over when the one person who gets it goes on holiday.
We are small. Are we too small?
Small teams get the most out of this, because there is nobody spare to soak up the busywork. Every hour it eats is an hour somebody needed. Run the calculator. If the number is smaller than a build would cost, I will tell you not to do it.
What does it cost?
The call and the plan cost nothing. You see the numbers in writing before money comes up. The build is priced per job, because a two-person office and a thirty-person shop are not the same work and a price range would help neither. That price is fixed in writing before anything starts. No hourly billing. No extra charges at the end. If the build costs more than the problem does, I will say so.
Do you need access to our data, and what stops it leaking?
Read-only to start, and more later if the job needs it. Three things keep it tight. I work inside your accounts instead of copying data to mine, so there is no second copy sitting somewhere. I take the smallest access that does the job, not full admin because it is easier. And you switch my access off when the job ends. I will sign whatever agreement your lawyer wants, theirs not mine.
What happens when you are gone and it breaks?
You own all of it, and it runs in your accounts, so there is nothing to be locked out of. You get notes written for whoever looks after it: plain words and screenshots, not a tech document. If it breaks you can call me, and most fixes are quick. The point is you choose to call, not have to.
Why is the audit free? What is the catch?
No catch, and here is why it works for me. Most of what I would put in a paid audit I can find in a few hours. The people who read the plan and see the numbers tend to want the build. The ones who do not were never going to buy, and I would rather learn that in three days than three weeks. Take the plan and do it yourself if you like. Plenty do.
What if we hate what you build?
We agree the job in writing first, so this does not happen at the end. You see working pieces as they land instead of one big reveal, so anything going the wrong way gets caught in week one. And if it is not running and doing what we agreed by the date, you do not pay the rest.
Can you work with our software, or do we have to switch?
You do not switch. Replacing software that works costs money, breaks habits, and is rarely needed. I work with what you already pay for: QuickBooks, Excel, Google Sheets, Shopify, Airtable, and most other apps. If one truly cannot be automated, I will tell you on the call instead of in week three.
How much of our time does this take?
A few hours in total, and most of it up front. The 30-minute call, then one session where somebody walks me through how the work really gets done, then the odd question after that. You hand over a problem. You do not manage a project.
Do you work with nonprofits?
Yes, and the work is the same. Donor reports, grant tracking and board packs are the same problem as invoices and stock, in different words. I will work with you on what fits your budget, and I will say plainly when it will not pay for itself. One thing worth checking before you assume there is no money: this kind of work often counts under capacity-building grants.
Book a 30-minute call
Tell me what gets done by hand right now. I read every one of these myself and reply within one business day.
Late means you do not owe the balance.
Book a call