Information security for software delivery

ISO/IEC 27001
ISO/IEC 27001 certified
The framing
Why a feature laundry list buys the vendor most willing to promise
We do not have an office in Saudi Arabia
We serve organisations there from our Riyadh office and our engineering hubs in Lahore and Karachi, travelling for the parts of an engagement that need to happen in person.
When an RFP is a hundred numbered features, every bidder replies with a hundred ticks, and the ticks say nothing — agreeing to a line item costs a sales team nothing at all. The comparison then collapses to price and confidence, and the bidder who wins is the one least troubled by what the features will actually take to build. The buyer has selected for willingness to promise, which is close to the opposite of engineering judgement.
The alternative is to write the RFP around the outcome and the constraints. State the business result the software must produce and how it will be measured; state the constraints that are genuinely fixed — the deadline and what drives it, the systems that must be integrated, the compliance obligations, the shape of the budget — and let bidders propose the path. The proposals will finally differ, because a path is a piece of thinking, and thinking is what you are buying. A bidder who cannot propose one without a feature list to echo has told you what they sell.
The clauses
The sections a custom-software RFP needs before it leaves procurement
Each row is a section the document should carry, with the test of whether a bid took it seriously.
The outcome and its measure
The business result the software exists to produce, the number that should move, and who owns that number today. Features may illustrate; they must not replace this.
What a good bid shows: A proposal argued from your outcome, not a quote per feature.
The constraints that are real
The fixed deadline and what drives it, the non-negotiable integrations, the compliance obligations, and the languages the product must serve — separated clearly from preferences that are open to challenge.
What a good bid shows: A plan built around the fixed points, and a challenge to at least one preference.
The current estate, disclosed
The systems the software must talk to, with versions and interface realities; where the data lives and what state it is in; who operates what. Estimating against a hidden estate is guessing, and you pay for the guess either way.
What a good bid shows: Integration effort priced against your named systems, with the risks called out.
Evaluation criteria that reveal engineering
Score a walkthrough of a past technical decision — what was chosen, what was rejected and why — and the seniority of the people actually assigned. Client logos and headcount slides score nothing.
What a good bid shows: A real decision narrated by the people who made it, not a reference list.
Commercial structure options
Invite bidders to propose the structure that fits the uncertainty — fixed scope where the problem is known, time and materials with a cap where it is not, or a paid discovery phase that prices the rest. Forcing a fixed price onto an unknown problem buys padding or disputes.
What a good bid shows: A structure argued from your uncertainty, with its trade-offs stated.
The interview
Four questions that show whether a software bidder builds or rents bodies
Put these to the shortlist in person. A body-shop can survive a document; it struggles in a conversation about decisions.
1 · Walk us through a technical decision you later reversed
A builder has reversals and can narrate one: the context, the wrong call, what the evidence was, what the reversal cost. A firm that has never been wrong has never been responsible for an outcome. Evidence the firm owns outcomes, not hours.
2 · Who will write our code, and who reviews what they write?
Named engineers, their seniority mix, and the review discipline between them. If the team will be "assembled at kickoff", the bid was priced on a bench you have never met. Names and a review structure you can hold the firm to.
3 · Tell us something in this RFP you would cut, and why
A builder reads your document critically and risks telling you so, because scope discipline is the job. A bidder who finds nothing to cut is agreeing with everything to lose nothing. A judgement made against your own document, in front of you.
4 · What do we receive besides the code?
Tests, deployment pipelines, environment documentation, an operational runbook, and the repository in your name from day one. A firm that delivers code alone is planning to be needed forever. A deliverables list your team could run the system from.
Our role
How we respond when a software RFP is written this way
- 01
We bid the outcome
Our proposals argue from the result the software must produce and challenge scope that does not serve it — because an RFP written around an outcome deserves a bid written the same way.
- 02
We name the team
The engineers who would do the work are named in the bid with their backgrounds, and the walkthrough of a past decision is given by the people who made it, not by a sales lead.
- 03
We price the structure to the uncertainty
Where the problem is well understood we quote it fixed; where it is not, we say so and propose the discovery that would make a fixed number honest, priced on its own.
The award is yours, made against your criteria; this page exists so the criteria reward whoever would actually engineer the work — and we are content to compete on that ground.
Common questions
What buyers ask
How is an RFP for custom software different from one for licensed software?+
A licensed-software RFP compares products that already exist, so a feature checklist works: the product either has the capability today or it does not, and you can verify it in a demo. A custom-software RFP buys something that does not exist yet, so the checklist verifies nothing — every bidder can promise every line. What you can evaluate is the team, its method, its judgement and its handover discipline, which is why the document should be built to surface those. Mixing the two — a feature checklist sent to development firms — produces bids that are all identical on paper and wildly different in reality.
Should a software RFP prescribe the technology stack or leave it to bidders?+
Prescribe only what you have a real reason to prescribe, and state the reason. Legitimate constraints exist: the platform your team can operate, an enterprise agreement already paid for, a compliance obligation, an estate standardised on one cloud. Those belong in the constraints section with their rationale, so bidders design within them. Beyond that, prescribing the stack converts the bid from an engineering proposal into a staffing quote, and it silently transfers design risk to you — if the prescribed stack turns out to fit badly, that was your decision, not the vendor's. Asking bidders to propose and justify a stack is also one more window into how they think.
What belongs in the current-estate section of a software RFP?+
Everything the new software must coexist with, at the level of honesty you would want if you were estimating the work yourself. Name the systems it will integrate with, their versions, and whether each exposes a usable interface or will need one built. Describe the data the software will read and write — where it lives, roughly how much there is, and how trustworthy it is, including the known problems. Say who operates the environments it will run in and what the deployment reality is. Where detail is commercially sensitive, provide it under NDA to the shortlist rather than omitting it. Every gap in this section returns as either a padded price, a mid-project change request, or both.
In this cluster
A working session on your RFP draft — what to reframe, what to disclose and what to ask, before it goes to any bidder including us.
Review your RFP draft with us