The decisions you cannot cheaply undo
Most of what a technology leader does day to day could wait a week. A handful of decisions cannot, because reversing them later costs a rebuild: the architecture, the platform, whether to build or buy, which agency you sign, who your first technical hire is, and what you do with the data you are collecting.
Those are the decisions I am there for. Not to run a standup.
Who this is for
- A founder whose product has outgrown the developer who started it
- A family business modernising operations with no one internally who can judge a vendor’s proposal
- A company mid-way through a build that has stopped feeling under control
- An investor or board that wants a technical read before committing
What I actually do
Set the direction, in writing
A technology strategy a non-technical board can read and act on: what we are building, what we are buying, what we are deliberately not doing this year, and what each choice commits you to.
Hold vendors and teams to account
Reviewing scopes before you sign them, reading what is delivered against what was promised, and asking the questions a client without a technical counterpart never gets to ask. This alone usually pays for the engagement.
Architecture and stack selection
Chosen for the team you have and the one you can realistically hire in this market — not for what is fashionable. The most elegant stack is worthless if nobody in the country wants to maintain it.
Technical due diligence
Before an acquisition, an investment or a large contract: what actually exists, what is held together with tape, what the real cost of ownership looks like, and which risks are worth pricing in.
Hiring the first few
Writing the brief, sitting in the interviews, and being honest about whether you need a senior engineer, a delivery manager or nobody at all yet.
How it works in practice
A fixed number of days a month, agreed up front. Enough presence to be useful and to know what is going on; not so much that you are paying an executive salary for oversight. Most engagements settle at two to four days a month, with more at the start and less once the direction is set.
I am reachable between those days. Decisions rarely wait for a scheduled meeting.
What you get that a full-time hire would not give you yet
Thirty years of seeing which of these decisions age badly. A CTO hired at the wrong stage spends their first year building the team they are used to. At this stage you usually need judgment more than headcount, and judgment does not have to be full time.
What I will tell you
If your problem is a process problem rather than a technology problem, I will say so. If a much cheaper off-the-shelf product does ninety per cent of what you asked me to build, I will point at it. That is the job.
Related reading
UAE e-invoicing is a systems project, not an accounting one is a worked example of the kind of decision that lands on a CTO rather than on finance.