Enterprise delivery credentials
NCA registration
Registered cybersecurity service provider

ISO/IEC 27001
ISO/IEC 27001 certified
The arithmetic
Why the minimum team is larger than it looks
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.
Continuous coverage means a position staffed at every hour of every day. A year has roughly 8,760 hours; one analyst working a normal schedule with leave and training covers somewhere near 1,700 of them. Covering one seat continuously therefore needs about five people before anyone has been ill, resigned or sent on a course, and a single seat is not a team — it is one person with nobody to escalate to at three in the morning.
The realistic floor for an in-house capability with genuine round-the-clock response is therefore eight to twelve people across tiers, plus a detection engineer who writes and tunes the rules, plus someone who owns the platform. Below that number an organisation has business-hours monitoring with an on-call rota, which is a legitimate choice and should be described as such rather than as a security operations centre.
The second constraint is retention. Tier-one monitoring is repetitive work that competent analysts leave within eighteen months, which means the capability needs a path upward for the people in it or a permanent recruitment pipeline. Organisations that build a SOC and then lose the two people who understood the tuning end up with an expensive platform producing alerts nobody trusts, which is worse than the business-hours arrangement they replaced.
Side by side
Three options, honestly costed
The middle option is the one most organisations end up in and the one least often named in a business case, which is why it is listed here as a choice rather than as a compromise.
| Option | What it takes to run | Where the cost sits | What you keep | Main failure mode | Best for |
|---|---|---|---|---|---|
| In-house SOC | Eight to twelve analysts across tiers, a detection engineer, a platform owner, and a recruitment pipeline. | Overwhelmingly salaries. The platform licence is usually under a fifth of the total. | Everything. The detection logic, the context, the institutional memory of your own estate. | Attrition. Two departures can hollow out the capability while the cost stays the same. | Organisations large enough to offer analysts a career path, and those whose estate cannot be described to an outsider. |
| Business hours plus on-call | Two to four analysts, a documented escalation path, and an agreed definition of what wakes someone up. | Modest, and predictable. The platform choice matters more here because there is less staff time to compensate for it. | Most of the context, and the ability to move to either of the other two later. | The gap is real: an intrusion that begins on a Thursday evening has until Sunday morning. | Most mid-sized organisations, as an honest starting position rather than a stated ambition. |
| Managed detection | A provider's analysts, plus one or two of your own people who own the relationship and the response. | A subscription, usually per endpoint or per volume. Predictable, and it rises with the estate. | The response decisions. The detection engineering and much of the context sit with the provider. | Alerts arriving without the context to act on them, and a slow drift into treating the provider as the security function. | Organisations that need continuous coverage sooner than they can hire for it, and that keep response in-house. |
What goes wrong
Four failures that are not about the choice
- 01
Buying coverage without response authority
A provider detects at two in the morning and telephones someone who is not permitted to isolate a host. Detection without a pre-agreed authority to act converts a fast alert into a slow one, and this is settled in the contract rather than during the incident.
- 02
Log coverage decided by cost
Ingestion is priced by volume, so the sources that produce the most data get dropped first — which is usually endpoint and authentication logging, the two that matter most for detecting lateral movement. Decide coverage on detection value, then negotiate the volume.
- 03
Detections that were never tuned to the estate
A default rule set produces a volume of false positives that trains everyone to ignore it within a month. Tuning is ongoing work by someone who knows what normal looks like here, and it is the first thing dropped when the team is short.
- 04
No rehearsal until the first real incident
The runbook is written, filed and never executed. A tabletop exercise twice a year finds the missing phone number, the account that expired and the decision nobody is authorised to make, at a cost of an afternoon.
Where we sit
What we do here
This page compares options rather than selling one, and part of being useful is being explicit about which of them we are and are not in a position to provide.
What we do
- Build the three-year comparison on your own event volume, retention requirement and coverage expectation.
- Design the detection use cases, the log pipeline and the response runbooks, whichever option is chosen.
- Run tabletop exercises against the runbook, which is where the gaps in it become visible.
Operating managed security services in Saudi Arabia is a licensed activity, and the licensing position is set by the National Cybersecurity Authority. Confirm a prospective provider's current status directly with the Authority's own published register rather than relying on a claim in a proposal.
Primary sourceCommon questions
What buyers ask
Can we start with managed detection and bring it in-house later?+
Yes, and it is a sound sequence, but the transferability has to be built in from the start rather than discovered at the end. Ask for the detection logic to be documented and portable, keep the raw logs in storage you own rather than only in the provider's platform, and retain at least one of your own people close enough to the tuning to reproduce it. Without those three, the migration is a rebuild.
Does data residency force an in-house SOC?+
Not by itself. What matters is where log data is stored and processed, and several providers operate infrastructure inside the Kingdom for exactly this reason. What the obligation is for your specific data is a determination for your legal and compliance function against the regulator's published rules — not something a supplier should assert on your behalf, in either direction.
What is the smallest arrangement that is genuinely worth having?+
Endpoint detection deployed everywhere with alerts going somewhere a named person reads daily, authentication logging retained for at least ninety days, and a written escalation path with a phone number that has been tested. That is not a security operations centre and should not be called one, but it detects the majority of what actually happens to organisations of that size, and it costs a fraction of the alternatives.
In this cluster
A three-year comparison built on your own event volume, retention requirement and coverage expectation, rather than on a generic model.
Model the three options on your numbers