The same order, from a workstation to a project
The method does not change with the size of the job. It holds for replacing a computer as much as for rebuilding a plant’s network, and it has four moments.
1. First you take the picture, without touching anything
The first step is not taking control: it is understanding what is there. Servers, workstations, network, access, backups, contracts in force and renewal dates — and above all how things depend on each other.
It looks like a bureaucratic step and it is the one that avoids the most expensive mistake in this trade: discovering a dependency on the day something gets switched off. In most companies nobody can say precisely what stops if a particular server is turned off, and that knowledge does not come out of a sales conversation.
What comes out of it is a document that is yours from day one, even if you then decide not to go ahead.
2. Then you decide with the numbers in front of you
Priorities, costs, timescales and an order of execution. The number everything depends on is almost always the same and almost nobody has calculated it: what an hour of downtime costs your company. From there come the two parameters that guide every infrastructure decision — how much work you can afford to redo, and how many hours you can be down.
Without that number, every proposal is a matter of taste: usually oversized where it does not matter and insufficient where it does.
At this stage we sometimes conclude that nothing needs doing, or that far less needs doing than had been proposed. We say so.
3. Then the work runs in blocks
Never everything at once. Each block has a useful result at the end, an agreed window and a way back if something does not go as expected.
In companies with production or shift work this is not an organisational nicety: it is the only way of working without stopping anything. A project that asks the company to stand still in order to be delivered is a badly designed project.
The same principle applies when changing supplier, in layers: first monitoring and helpdesk, which sit alongside without switching anything off; then security and backups; last the systems that, if handled badly, stop the company.
4. Then it stays written down, and it stays yours
Every piece of work leaves a trace you can consult: what happened, what was done, how long it took. Configurations, diagrams and credentials are documented and can be handed over at the end of the relationship too.
This is not a courtesy: it is your company’s technical asset. A supplier who keeps the map of your systems to themselves is not protecting you, they are tying you in — and that sentence applies when the supplier is us.
The three questions we suggest asking anyone
We put them in writing because they hold up when aimed at us too.
How many people actually answer the phone? Not how many appear on the website. The answer comes in two seconds or it does not come.
What happens when my contact is away? If it is “I will get them to call you back”, there is no cover. If it is “anyone picks it up, because your configuration is documented”, there is.
Is the documentation of my systems mine? A vague answer is already an answer.
The first step
An assessment visit: a consultant looks at the state of the infrastructure, says what makes sense to do, and leaves a written picture with the priorities. Free and without obligation, even if you then decide to change nothing.