Why the generalist software house loses

Ask a software company in Egypt what it does and the answer usually arrives as a list: web, mobile, ERP, e-commerce, custom systems, and lately AI. The list is honest. It is also why the last three proposals were compared against two other firms and decided on the number at the bottom of the page. A buyer reading that list learns what you are capable of operating. He learns nothing about what you have already solved.

What the general position costs

The first cost is that the buyer has to do the imagining. A finance manager who needs to close the month in four days instead of fifteen reads "custom software development" and has to build the bridge himself, from your capability to his ledger, his branches, his tax filing. Most buyers will not do that work. They contact the company whose page names their problem, and the general firm never learns it was in the running.

The second cost is arithmetic. When three vendors present the same undifferentiated capability, the proposals are compared on the only two fields that differ: total price and delivery date. Nothing else on the page gives the buyer a reason to prefer one over another, so the lowest number wins, and the winner discovers the scope was larger than the price. This is a failure of position long before it is a failure of pricing.

The third cost is slower and heavier. Domain knowledge compounds only when the domain repeats. The second contracting company you build for should consume fewer hours than the first: you already know how retention on a subcontractor works, how a progress claim is structured, why the site engineer will not fill in a form with fourteen fields. If every project sits in a new sector, that discount never arrives. Six years pass and the margin looks the same as it did in year one.

The position is already in the delivery record

The first instinct when narrowing is to invent: pick a sector that is growing, write a new line, announce it. That produces a claim the company cannot support and quietly abandons within a year. The position you can hold is the one already visible in what you have delivered.

Take the last fifteen to twenty projects and answer four questions about each of them, with real numbers:

1. Which were profitable measured against hours actually worked, not hours invoiced?
2. Which were delivered inside the estimate, or ahead of it?
3. Which did you understand without being taught, where discovery was one call rather than a month of meetings?
4. Which produced a referral you did not have to ask for?

Then look for the overlap. In most companies three or four projects sit in all four columns, and they are frequently not the projects the team talks about. The technically interesting work and the profitable work are rarely the same work, and the position follows the second.

A position taken from the record can be shown. A position that was invented has to be argued every time somebody questions it.

The exercise fails in one predictable way: sampling from ambition rather than from delivery. If the answer to "which sector did we understand immediately" is a sector you have delivered in once, that is a hypothesis, not a position. Keep it on the list and test it with the next two projects.

The axes of specialisation are not equivalent

Specialisation is not a single decision. There are four common axes, and they differ in how much they compound and how easily a competitor copies them.

Sector is the most durable axis and the hardest to copy, because it is bought with years rather than with hiring. System type is the most portable: an accounting core built for one sector moves into the next with weeks of work rather than months. Buyer size is rarely a public headline, but it governs everything operational — the contract, the length of the cycle, whether there is a procurement department at all. Technology is the weakest public position, and unfortunately the one most companies choose, because it is the easiest to write.

The positions that hold are usually two axes crossed. "Accounting and inventory systems for multi-branch retailers in Egypt" is a sector crossed with a system type: narrow enough to be believed, wide enough to feed a company of fifteen people. It also has an anchor a general firm cannot fake. Anyone selling accounting systems in Egypt lives inside the Tax Authority's e-invoicing requirements, and a buyer can tell within a few minutes of conversation whether you actually live there.

Narrowing does not mean less work

The fear is legitimate and deserves a direct answer rather than reassurance. Yes, narrowing reduces the number of inquiries. It raises the share of inquiries that reach a scoped project, and it raises the price you can hold, because the buyer is no longer comparing you with two firms who might equally be building a school website.

There are two real risks. The first is choosing a segment too small to feed the company, and the test for it is arithmetic. Name the organisations in Egypt, or in the Gulf if you sell there, that fit the position and have budget for a custom system. If you cannot reach forty or fifty names, the axis is narrower than it should be and needs to widen by one step: to an adjacent sector, or from the system type out to the operating problem around it.

The second risk is concentration. A company whose entire pipeline sits in one sector moves when that sector moves, and sectors in this region move sharply with the exchange rate, with import rules, with a single regulatory decision. The remedy is sequence rather than hedging. Establish the first position until it produces referrals on its own, then add a second, adjacent one. Adding the second before the first is established simply recreates the general firm with more words.

The public position and the private capability

Narrowing is a decision about what the market is told. It is not a decision about what the company is permitted to build. Confusing the two is why most positioning attempts collapse at the first inquiry from outside the stated field.

The public position is one sentence, and everything the market touches repeats it: the site, the proposal, the outbound message, the improvised answer at a conference. The private capability is whatever the team can actually deliver, and it is always broader. A project referred by a good client from outside the position can still be accepted on your normal terms; there is no sense in refusing good work on doctrinal grounds.

The discipline lies in what follows acceptance. Off-position work is not written up as a case study, does not enter the outbound list, is not discounted "for the portfolio", and is not added to the services page. In practice a company can hold a clear public position while a third of its revenue comes from adjacent work. What matters is that the adjacent third is not what the market hears about, because the market will remember one thing about you and no more.

Rewriting the site, the portfolio and the pitch

The position is not real until the material says it. Four changes, in order:

• The first line of the homepage. It names the buyer and the system, not the ambition. "Accounting and inventory systems for multi-branch retailers in Egypt, in production and under our support" tells a buyer whether to keep reading. "We deliver innovative software solutions" tells him nothing.
• The services page. Cut it to three or four entries that sit on the position or directly support it. Eleven services is a statement that none of them is a speciality.
• The portfolio. On-position work first, and every entry states the operating problem, the constraint, what runs today and since when. That is what turns a record into proof a buyer can check.
• The sentence everyone says. Ask five people in the company what it does and you should hear five wordings of one sentence. If you hear five different sentences, the position exists only in a document.

Search follows the same logic. Pages that answer the specific operating question — how stock transfers between branches are handled, what e-invoicing compliance actually requires — are reachable by the person who has that problem today. A page aimed at "software company in Egypt" competes with everyone and convinces no one.

A position has to be falsifiable

Any company can write that it specialises in something; the sentence costs nothing. The position becomes real at the moment somebody can check it: name the systems, say what each one does, say how long it has been running and at what scale, and be willing to put a client on a call. If three delivered examples on the position do not exist, what you have is a direction rather than a position, and the honest move is to say it in those terms while the record is being built. Softwiro's own portfolio names the systems it has built, most of them linked to an address a reader can open, for that reason.

If you are trying to work out what your delivery record already supports, or want a project scoped against it, contact us. The narrowing itself takes one session with the delivery record, and a year of holding to it.