Companies rarely buy an operations system because the old one broke. They buy it because the business outgrew the number of things one person can hold in their head.
The real incumbent is not software
A custom operations system almost never replaces a competing product. It replaces spreadsheets that disagree, a messaging thread standing in for a workflow, and institutional knowledge held by whoever has been there longest. That is why the migration is a change management problem before it is a technical one.
This matters for scoping. Replacing a known system means a specification already exists, written by the thing you are replacing. Replacing an undocumented process means discovery is the project, and skipping it means building somebody's guess about how the business works.
Where the pain actually starts
The trigger is usually one of four, and they arrive in a predictable order as a company grows.
- Two sources of truth. Sales quotes from one sheet, the warehouse works from another, and the reconciliation happens in an argument.
- A person who cannot go on holiday. One individual is the integration layer between departments. That is a risk with a name on it.
- Retyping. The same order is entered three times in three systems, and the error rate is exactly what you would expect.
- Questions nobody can answer quickly. How many units are actually available. What did this client agree to. Which orders are late and why.
What gets built, and in what order
| Order | What | Why first |
|---|---|---|
| 1 | The one record everything hangs off | Unit, order, client. Get this wrong and every later decision inherits it |
| 2 | State and who may change it | Available, reserved, sold. Ambiguity here is where the money leaks |
| 3 | The daily workflow | The screens the team touches every hour, designed for speed not completeness |
| 4 | Reporting | Only once the underlying data is trustworthy, never before |
| 5 | Integrations | Accounting, messaging, whatever already works and should keep working |
Reporting sitting at position four surprises people, because it is usually the reason the project got funded. A dashboard over data nobody trusts just distributes the mistrust faster.
Two we have built
Travelholic runs more than a hundred furnished apartments across Cairo. We built the operating system behind that: one modular platform covering unit state and the back office, with an AI layer inside it rather than a booking tool bolted onto spreadsheets.
Egyptian Marble & Granite runs a marble manufacturing and retail network. The system we built runs the floor, the orders and the people, which in practice means the shop floor and the sales side finally reading from the same record.
Different industries, the same underlying shape: one trustworthy record, explicit state, and a workflow built for the people who use it every day.
When not to build one
If your process fits a well-supported off-the-shelf product without meaningful compromise, buy it. Custom earns its cost when the operation itself is the unusual part, when the workflow is a competitive advantage rather than an inconvenience, or when what you need to connect simply is not connectable any other way.
The honest test: write down the three things your business does differently from a direct competitor. If none of them touch the system you are considering building, buy instead.