Hong Kong
We build the software a Hong Kong business actually runs on — the internal tools, portals and operational systems no vendor sells — and hand the whole thing over: code, infrastructure and documentation, in your own accounts.
The shape of the problem
Nearly every Hong Kong business we meet has one. It started as a convenience and became the operation.
It has a guardian — the person who knows which tab is authoritative and which formula must not be touched. It is emailed around as v7-final-FINAL. Two people open it at once and one of them loses an afternoon. When that person is on leave, a part of the business quietly stops.
A spreadsheet is the fastest way in the world to model a process. It is one of the slowest ways to run one, because it has no rules: nothing stops a date going in the wrong column, nothing records who changed what, and nothing tells you when a step did not happen.
That is not a spreadsheet problem. It is a sign that a real process outgrew the tool holding it, and that no product on the market fits the process closely enough to replace it.
Build or buy
Worth saying first, because most consultancies will not: custom software is the wrong choice more often than it is the right one.
Accounting, payroll, email, storage, CRM for a conventional sales process — these are solved problems with mature products, and anything we built would be a worse, lonelier version of software a thousand companies have already stress-tested. If a product genuinely fits, buy it. We will tell you so at no charge, and that is frequently how a first conversation ends.
Custom earns its cost in a narrower place: the processes that are genuinely yours. The ones where every off-the-shelf tool you have tried needed a spreadsheet bolted to the side of it to be usable. Those are the processes that make your business different from its competitors, which is exactly why no vendor has built for them.
What gets built
Different industries, the same underlying shape: something operational, used daily by a small number of people who cannot afford it to be wrong.
The screen your operations team lives in all day. Built around the decisions they actually make, in the vocabulary they actually use.
Somewhere for the people outside your company to submit, check and approve things themselves, instead of emailing someone who retypes it.
Pricing rules that are complicated for good reasons, availability that depends on more than a calendar, and a quote that cannot be sent with a stale number in it.
Where things are, who has them, what state they are in, and what happened to them — recorded as it happens rather than reconstructed at month end.
The four or five numbers you actually manage by, computed once, from one source, so two people asking the same question get the same answer.
The data model, permissions, validation and error handling. This is where the value of custom software actually lives, and it is what gets built first.
Hong Kong businesses tend to arrive with a particular set of constraints, and a system that ignores them needs rework within a year.
Built for here
Not a template with a Hong Kong address on the contact page.
Cost and timeline
Scope drives the number, so anyone quoting a figure before understanding your process is guessing. What we can be precise about is the shape.
From first conversation to a working first version running on real data — one workflow, end to end. Not a prototype and not a specification.
Adding the eleventh user does not cost more than the tenth, because there is no licence. Your costs stop scaling with your headcount.
Hosting, storage, third-party APIs and the maintenance a live system needs. It is the number people forget to ask about, and it is usually the smaller half.
We widen into neighbouring processes only once the first one has proven itself in production. It is the cheapest way to be wrong about something.
Handover
Every engagement ends with the repositories, the infrastructure-as-code and the documentation transferred into your own accounts, and your team trained on how it works. There is no proprietary runtime, no per-seat licence, and nothing that prevents another developer picking the work up after us.
This is the part worth checking with any vendor you talk to, here or anywhere. Ask where the code will live, whose cloud account runs it, and what happens to your data if you stop paying them. The answers vary more than you would expect.
Questions
It depends on the scope, but the honest shape is this: a first working version covering one workflow end to end is a matter of weeks, not quarters, and it is priced as a project rather than per seat. The number that surprises people is the running cost, so we quote that separately and up front — it is usually the smaller half.
About four weeks from the first conversation to a working first version running on real data. That is one workflow built end to end, not a specification or a prototype. Widening into neighbouring processes happens after that has proven itself.
Yes, literally. The repositories, the infrastructure definitions and the documentation are transferred into your own accounts at handover. There is no proprietary runtime, no per-seat licence and nothing that stops another developer picking the work up.
Often, yes — and we will say so. Off-the-shelf software fits standard processes well and unusual ones badly. If a product genuinely covers what you need, buying it is cheaper than anything we could build.
That is usually most of the job. Very little of this work starts from nothing: it connects to the accounting, CRM, e-commerce or warehouse systems already in place, and the integration is designed before the interface is.
Related
When the problem is not a missing tool but five tools that disagree with each other.
When the work itself is the problem: the intake, approvals, handoffs and chasing that live between the tools.
The four kinds of work in full, including systems consultancy — the audit and the build-versus-buy calls.
Understand before building, ship something small into production, then hand it over.
Describe the process or the spreadsheet costing you the most, and we will give you a straight read on whether it is worth building — including if the answer is no.