Most small businesses are carrying two or three jobs that a piece of software should be doing. They rarely get fixed, because nobody on the payroll has the time and nobody wants to buy a platform to solve one problem. I build the one thing.
There is no product that does the specific thing your business does, because your business is the only one that does it that way. The platforms that come closest cost more per year than building the thing once, and they still need somebody to work around the parts that don't fit.
So it stays manual, and the cost never appears on any line of the accounts. It appears as somebody's Tuesday.
The rule I work to: buy when a product fits the way you already work; build when the way you work is the thing that makes you money. If you are on the buy side of that line I will say so on the call, and you can go and buy it.
These run on invented data, and they are the standard I build to rather than sketches of it. Each one answers some of the symptoms above.
We find the problem — what the task is, who does it, how long it takes. Then we settle what gets built, what it costs and the terms, in that same conversation. If something off the shelf already does it, I will tell you.
Usually starts the same week. Most of these are two or three days of work and a larger one runs to about a week — not because they are trivial, but because one tool is a much smaller thing than a product. You see it working as it goes.
Running on accounts in your name, with commented source and a note on how it works, so another developer could pick it up. I stay reachable, weekends included. Support is monthly and never a condition of the build.
Their reps were cold-calling car dealerships with no way of knowing which ones already sold extended service contracts, so most calls reached somebody who was never going to buy. The symptom at the top of this page — sales people spending the morning finding someone to call — is this one.
It now searches a territory, reads each dealership's own website, works out whether coverage is offered, and hands the rep a ranked call list with the evidence, the contact and a drafted opening email attached. Call outcomes are shared, so two reps never work the same dealership twice.
It is also the pattern I use on my own prospecting, which is the most honest recommendation I can give it: the tool that finds my clients is one I built with the same parts.
I scope the work, write the code, deploy it and answer the phone afterwards. There is no account manager between us and no handoff once the contract is signed. You explain your business once, to the person who is going to build for it.
Agreed before anything is built. Scope changes are repriced openly rather than absorbed quietly.
Source, hosting, domains and data sit in accounts in your name. Walking away costs you nothing but the work already done.
Most jobs are two or three days; a larger one runs to about a week. A scoped project has a date on it, not a roadmap.
Send something on a Friday evening and you will have an answer before Monday. It is most of why these finish in days.
The honest limit of working alone. If I am full I will say so rather than take a deposit and put you behind two other builds.
Pick the cheapest job on your list and we do that one first. A small tool that works tells you what the next one is worth far better than a proposal does.
Most of these cost nothing to about $20 a month to run, paid directly by you because the accounts are yours. No licence and no seat fees.
Anything that moves money, and anything in payroll, tax, medical or legal. Being wrong there costs you more than the job is worth.
By making it small enough that it can't be one. We pick the cheapest thing on your list, agree a fixed price for that one thing, and build it in a few days. If it works you know exactly what the next one is worth. If it doesn't, you are out one small invoice rather than a project.
The money pit happens when scope is open and billing is hourly. Neither is true here: the price is agreed before I start and there is no hourly rate to run up.
Usually nothing, or close to it. Most of these run on free tiers — expect zero to about twenty dollars a month, and you pay that directly because the accounts are in your name. There is no licence, no seat fee and no subscription to me.
Often, yes, and I will tell you so on the call. Buy when a product fits the way you already work. Build when the way you work is the thing that makes you money and no product matches it.
Sending you to a product I don't sell costs me one job and is the reason you would call me again.
The code lives in a repository in your name from the first day, not mine, along with a written note explaining how it works — written for a developer who has never spoken to me. Hosting and data are in your accounts too.
It is a fair question and the answer should not be "trust me." It should be that you could hand the whole thing to somebody else tomorrow.
Probably not. The jobs worth fixing here are small by definition — one task, a few days of work, two thousand dollars at the low end. If you have five people and one of them loses a day a week to retyping, that is exactly the shape of thing this is for.
Too small would be a job that saves an hour a month. I will say so.
Nothing on this page is a contract, and the figures on it are not a quote. Work happens under a short written agreement signed before anything starts.
The hours, build times and price bands in the panel above are indicative and generated from typical figures, not from your business. They are not an offer, a quote, or a commitment to a price. Your actual scope and price are agreed in writing after a conversation.
Everything in the worked examples runs on invented data. The companies, figures, jobs and quotes in them are fictional and are there to show how a tool behaves, not to represent real clients or real results.
Every engagement runs on a written agreement covering scope, fixed price, payment schedule, what counts as finished, and what happens if either of us wants to stop. Ownership of the delivered work transfers to you on final payment.
Scope changes are repriced in writing rather than absorbed quietly.
Anything you tell me about how your business runs stays between us, and I am glad to sign a reasonable NDA. I do not name clients publicly without written permission, which is why the project on this page is unnamed.
Nothing. There is no analytics, no tracking pixel and no cookie set by this page. The checker above runs entirely in your browser and nothing you tick is sent anywhere or stored.
Booking a call goes through Cal.com, which collects your name, email and chosen time under its own privacy policy. Email reaches my inbox and stays there.
I will tell you honestly whether I think a job is worth doing, and the panel above will do arithmetic on figures you supply. Neither is a promise that your business will save a particular amount of time or money. What I commit to is delivering the agreed software, working, at the agreed price.
Software can fail. The written agreement caps what I am liable for at the fees paid for the work, which is the normal arrangement for a firm this size and the reason I turn down jobs where being wrong would cost you far more than the job is worth — anything moving money, and anything in payroll, tax, medical or legal.
Kichiro is a sole proprietorship based in Seattle, Washington, operated by Griffin Saito-Hardie. This section describes how I work; it is a plain-language summary and not legal advice. The agreement we both sign is the document that governs the work.
Tell me what the job is and roughly how often somebody does it. That is usually enough to say whether it is worth building and what it would cost.
Book a call
Or email Kichirodev@gmail.com
Usually a reply the same day, weekends included.