
Managed SOC Onboarding: What Happens Before Go-Live
Managed SOC coverage does not start on the signature date. The onboarding phases, who owns each log source, and the go-live test that proves detection works.
The short answer
Managed SOC onboarding is the work that turns a signed contract into real monitoring: agree the in-scope systems, connect each log source and prove events arrive, write down what normal looks like, tune detections, and fix who can act on an alert. Treat coverage as live only after a go-live test you have witnessed, not on the signature date.
Who this is for
This is for a CTO, head of security or operations lead at a casino, sportsbook, crypto exchange, bank or payments firm who has chosen, or is about to choose, a managed SOC and wants to know what the weeks after signing should look like. It sits between two other decisions. Comparing offers is covered in what a managed SOC quote must say, and the contract clause that decides when the response clock starts is in where the SLA clock starts. This article covers the stretch in the middle, when the SOC exists on paper and not yet in practice.
Why the signature date is the wrong start
A SOC does not watch what it cannot see. On the day a contract is signed, no log source is connected, no baseline exists and nobody has written down who answers the phone at three in the morning. Three things follow from that.
- Coverage starts at zero and grows. Every system you listed in the quote has to be connected, parsed and checked. Until then, an incident on that system is invisible to the provider.
- Detections are generic until they are tuned. Out-of-the-box rules produce either noise or silence on a platform they have not seen. Tuning needs time and your input.
- Response authority is a document, not a default. Unless the actions the SOC may take are written down and approved, the first real alert becomes a negotiation.
Most disputes about "the SOC missed it" trace back to a source that was never connected, a rule that was never tuned, or an action nobody had authorised. All three are onboarding gaps. They are cheap to close before go-live and expensive to discover during an incident.

The onboarding phases
The order matters more than the timings, and timings depend on your estate and the provider, so none are given here. Ask the provider for their own plan and hold them to its sequence.
| Phase | What happens | What you must provide | Evidence that it is done |
|---|---|---|---|
| 1. Scope confirmation | Quote scope is turned into a named list of systems, sources and hours | The real inventory, including what the quote did not mention | A signed source list that matches your own asset records |
| 2. Source connection | Agents, forwarders or API collectors are set up for each source | Access, network paths and an owner for each system | Events arriving from every listed source |
| 3. Quality check | Events are checked for completeness, timestamps and parsing | Confirmation of what fields and time zones are expected | A per-source note on gaps, lag and fields not captured |
| 4. Baseline | The SOC learns normal activity for your business | Shift patterns, release windows, campaign periods, admin habits | A written baseline you have read and corrected |
| 5. Detection tuning | Rules are enabled, adjusted and retired against that baseline | Feedback on early alerts: expected, noise or real | A list of active detections with owners for each |
| 6. Response authority | The SOC's allowed actions and contacts are written down | Named approvers, out-of-hours numbers and a default when nobody answers | A signed action table and contact list |
| 7. Go-live test | Harmless test actions prove detection and escalation work | Someone to perform the actions and witness the alerts | A test log showing action, alert and the person who received it |
| 8. Handover to run | Reporting, review meetings and change process begin | A named service owner on both sides | The first coverage review in the calendar |
Phases 3 and 6 are the ones most often skipped. Skipping the quality check means a source can be "connected" while sending half its events. Skipping response authority means the SOC can detect everything and do nothing.
Source by source: who owns what
Every log source needs a human on your side before the SOC touches it. The common failure is a source with no owner: nobody can approve the network path, nobody knows why events stopped, nobody answers when the SOC asks whether an administrator login is normal.
For each in-scope system, write down three lines before onboarding starts:
- Owner. The person who can grant access, change configuration and answer questions about normal use.
- Dependency. Whether the logs come from a third party, such as a hosted cashier, KYC provider or cloud account, and whether the contract lets them be exported.
- Sensitivity. What the logs contain that should not leave your control, such as player identifiers or card data, and how the provider will treat it.
If the system is run by a vendor, ask early whether log export is possible and at what cost. A source you cannot export is a gap you should know about now, not after an incident. The same discipline applies to the sector-specific choices in managed SOC for iGaming, where log sources and response boundaries are fixed before onboarding.
The baseline: teach the SOC your normal
A baseline is a short written description of what normal looks like for your business. It is not a statistical model you need to build, but it must exist, and you must read it.
Cover at least these points:
- Working patterns. Who is on shift when, from where, with which devices.
- Planned change. When releases, migrations and maintenance happen, and how the SOC will be told.
- Business peaks. Events, promotions or trading hours that legitimately change traffic and logins.
- Privileged access. Which accounts do administrative work, and what that work normally looks like.
- Known oddities. Shared accounts, legacy systems and scheduled jobs that look suspicious but are not.
The fifth item saves the most noise. An honest list of oddities lets the SOC suppress what is known and focus on what is not. A baseline also needs a change route: when your business changes, someone must tell the SOC, or the baseline decays into a source of false alarms.
Response authority: write it before you need it
Detection without authority is a notification service. Before go-live, agree a table in three tiers: what the SOC may do on its own, what needs your named approver, and what only you decide.
- SOC may act alone: low-impact, reversible steps, such as forcing a session to re-authenticate or isolating a single test host if you agree.
- Needs approval: actions that affect customers or revenue, such as blocking an account, disabling an integration or changing a firewall rule.
- Your decision only: notifying customers, regulators or law enforcement, and anything that suspends trading or payments.
Add the approver, a backup, how they will be reached and what happens when nobody answers. That last line is the one that matters at night. The structure is the same as the one used in incident response retainers, because the two services meet at the same point: the first hour of an incident.
The go-live test
A go-live test replaces trust with evidence. For each in-scope source, agree a harmless action you will perform and the alert you expect. Examples include a failed-then-successful login on a test account, creation of a dummy administrator in a non-production tenant, or a benign file added to a monitored path. Do not use anything that could harm production or resemble a real attack technique beyond what the test needs.
Record four things per test: the action and its time, the alert that appeared, who received it and how long the whole path took end to end. If a source produces no alert, that is a finding, not a failure of the test. It is the reason the test exists.
Agree in advance what counts as "live". A good definition is that every in-scope source has passed a witnessed test, the response authority table is signed and the contact list has been exercised once. Only then should the full monthly fee, and any response-clock commitment, begin.
What to put in the quote
Because onboarding sits between the quote and the first alert, it belongs in the quote. Ask for:
- The provider's onboarding plan, in phase order, with a named lead.
- What you are charged, and when, before go-live.
- Who is responsible for each source connection, on both sides.
- The go-live test and the definition of "live".
- A route and price for adding sources later.
- A way to be told when a source stops sending events.
If the quote says only "onboarding included" and nothing else, treat it as an unanswered question.
After go-live: keep the picture current
Coverage drifts. Systems are added, vendors change their log format, a collector is upgraded and quietly stops. Ask for a recurring coverage review that compares what is onboarded against what you actually run, plus a notice when any source goes silent. Without these, the green dashboard shows what was connected at kickoff and not what exists today.
Limitations
This article is not legal, regulatory or contractual advice, and it does not state what any licensing body, supervisor or standard requires of onboarding or monitoring. It uses no external statistics or case studies and gives no timings, because they vary by estate and provider. The phase table is a planning aid to adapt to your systems and contract. A completed onboarding shows that agreed sources are connected and tested at one point in time; it does not guarantee detection of every attack, and the provider can only see what is sent to it.
Decision and next step
If you can name, for every in-scope system, its owner, its log path and the test that proved it works, and you hold a signed table of what the SOC may do, you are ready to go live. If you cannot, the gap is in your onboarding, not in the SOC's rules. To have the phases, source list and go-live test written into a quote before you sign, request a coverage review with AnySec and we will map your systems to a source list, a baseline outline and a response-authority table.
Sources and review
The framing follows public guidance on logging and incident handling: NIST SP 800-92 Guide to Computer Security Log Management and NIST SP 800-61 Rev. 2 Computer Security Incident Handling Guide, read 2026-11-17; check current versions. AnySec prepared the phase table from the public structure of managed monitoring engagements; no AnySec-original statistics are used. Author: AnySec Engineering. Published 2026-11-17; reviewed by the AnySec security team.
Related reading
- What a managed SOC quote must say — the seven lines to compare before you sign.
- Choosing a managed SOC provider: where the SLA clock starts — the clause that defines when the response promise begins.
- Managed SOC for iGaming — fixing log sources and boundaries before onboarding.
To have onboarding specified up front, get a fixed quote.
Frequently asked questions
- What is managed SOC onboarding?
- Onboarding is the work between signing and the date monitoring is genuinely live: agreeing the in-scope systems, connecting log sources, confirming events arrive complete and on time, learning what normal looks like, tuning detections and writing down who the SOC may call and what it may do. Until that is finished, the SOC can see only part of your estate, however the contract describes it.
- How do I know when onboarding is finished?
- Agree a written go-live test in the quote. For each in-scope log source the provider shows that events arrive, that a harmless test action on your side produces the expected alert, and that the alert reaches a named person on your side. A completion email from the provider is not the same as a test you witnessed.
- Who needs to be involved from our side?
- A named owner for each log source, someone able to approve response actions out of hours, and whoever knows the normal working pattern of your business: shifts, release windows, campaign periods. If any of these roles is empty, onboarding stalls on the first question the SOC cannot answer from the logs alone.
- Should we pay the full fee during onboarding?
- Ask the quote to say what is charged before go-live. Some providers charge a separate integration fee, some start the monthly fee on signature, and some start it at go-live. None of these is wrong, but full fees with no stated onboarding plan means you pay for coverage that does not exist yet.
- Can we add log sources after go-live?
- Yes, and you will. New systems, acquisitions and migrations all change what needs monitoring. Ask how a new source is requested, whether it is priced, and how the provider will tell you when a source silently stops sending events.
Rather not learn this in production.
Talk to the engineers behind these write-ups — no sales script, a straight read on where you stand.
Get a fixed quote