Technology Consulting: What It Is, When You Need It, What It Should Cost

Technology consulting has a reputation problem, and it earned it. Too much of it produces a well-formatted document, a roadmap with three horizons, and an invoice — after which nothing changes. But the underlying need is real: most businesses face technology decisions worth hundreds of thousands of dirhams, made by people who reasonably don't know how to evaluate them. This is what the useful version looks like, and how to buy it without funding a report.

The four problems actually worth hiring for

1. "Should we build this or buy it?"

The highest-value question in the list, because getting it wrong costs six figures in either direction — building something you could have subscribed to, or subscribing to something that will never fit and rebuilding it two years later anyway. A week of proper analysis against a real shortlist is cheap insurance. Our framework for it is in no-code AI vs custom development.

2. "Our systems don't talk to each other."

The classic UAE SME situation: an accounting package, a CRM, a POS, three spreadsheets and a WhatsApp group, with staff manually moving data between them. The consulting job is mapping where time actually goes and identifying which two or three connections remove the most manual work. The answer is frequently automation between existing tools rather than new software, and it's usually cheaper than everyone feared.

3. "What should we do about AI?"

Currently the most common request, and the one where most money gets wasted. The useful engagement doesn't produce an "AI strategy" — it produces a shortlist of specific processes ranked by payback, a clear split between what to buy off the shelf and what's worth building, and a first project small enough to prove or disprove the case in weeks. Anyone selling an AI transformation programme before identifying one concrete process is selling you a document.

4. "Our software projects keep failing."

Late, over budget, or delivered and unused. Almost always a process problem rather than a technology one: unclear ownership, requirements that never stabilise, no acceptance criteria, no one empowered to say no. An outside review that names the actual cause is uncomfortable and usually worth more than the fee.

The engagement formats, and what each should cost

How to buy it without funding a report

1. 1Define the decision, not the topic. "Decide whether to replace our booking system by 31 August" beats "digital transformation strategy". A decision has an end date; a topic doesn't.
2. 2Cap it at four weeks. Anything longer stops being analysis and becomes a project with no deliverable.
3. 3Specify the deliverable in plain words. A recommendation, the options considered, the costs, and what to do first. Not a maturity model.
4. 4Ask who implements it. If they don't build, ask how their recommendations have survived implementation before. If they do build, ask them to state their bias openly — a partner who says "we'd benefit if you build, so here's what would make me wrong" is being straight with you.
5. 5Require one uncomfortable finding. Any honest review of a real business surfaces at least one. If the report agrees with everything you already believed, you paid for a mirror.

A good consulting engagement ends with a decision and a first task. A bad one ends with a document and a follow-up meeting.

When you don't need consulting at all

• You already know what to build. Skip to a scoped build with a fixed price. Don't pay someone to agree with you.
• The problem is one workflow. Describe it to a developer and get a number. A three-week strategy engagement for a two-week automation is a waste.
• The budget is small. If the whole project is AED 40,000, spending AED 20,000 deciding how to spend it is obviously wrong.
• The real problem is organisational. No consultant fixes an unclear owner or a leadership team that hasn't agreed on priorities. Technology will just make the disagreement faster.

How we do it

We consult as builders, which is both the advantage and the bias, and we'd rather name it than hide it. The advantage is that our recommendations have to survive us implementing them — a plan that doesn't work becomes our problem, not a footnote in a document. The bias is that we make money building, so we could be tempted to recommend building. We manage that in the most direct way available: we tell you when the answer is a subscription, an off-the-shelf product, or doing nothing yet, and we say it in the first conversation rather than after an invoice.

In practice most of our consulting is short and specific: a paid discovery that produces a scoped, costed plan for a real project — and if you then build it with us, that fee comes off the build. If you build it elsewhere, you still own the plan. We work across AI, automation and software for businesses in the UAE, and the case studies show where those recommendations ended up in production.