When part of an engineering organisation sits in India and part in Germany, Austria or Switzerland, incident response either benefits from the time offset or suffers from it. It benefits when coverage, roles and access are designed deliberately, so that the India-based team handles the early European morning and the overlap hours with clear authority. It suffers when nobody knows who may restart a service, who talks to customers, or whose phone rings at 03:00 in Stuttgart. The answer is a written incident model covering coverage windows, severity levels, roles, access rights, reporting duties and postmortems, agreed as part of the contract rather than improvised during the first outage. This guide sets out such a model.
The numbers: why incident design is a board topic
The threat level for DACH companies has not eased. Bitkom's 2025 economic protection study put the damage to the German economy from theft, espionage and sabotage at a record 289.2 billion EUR, with 87% of companies affected and around 70% of the damage, just over 202 billion EUR, caused by cyberattacks. One in three companies (34%) was hit by ransomware. The BSI reported in its 2025 situation report an average of 119 newly discovered vulnerabilities per day. Speed matters: IBM's Cost of a Data Breach Report 2025 measured an average breach lifecycle of 241 days to identify and contain, at a global average cost of 4.44 million USD. Regulation has followed. Germany's NIS2 implementation law took effect on 6 December 2025 and widened the circle of supervised entities from about 4,500 to roughly 29,500. An incident model across two time zones has to work under this scrutiny.
The coverage arithmetic
India Standard Time is UTC+5:30 throughout the year, with no daylight saving time. Central European Time is UTC+1 in winter and UTC+2 in summer. A standard Indian working day from 09:30 to 18:30 therefore corresponds to 05:00 to 14:00 in Germany during winter and 06:00 to 15:00 during summer. Combined with a German working day that begins around 08:00 or 09:00, business-hours coverage across both locations extends from early morning to late afternoon or evening in Central Europe, without anyone working at night.
This is not full follow-the-sun coverage. Nights and weekends in Europe remain uncovered unless someone is on call. But for many systems whose users are in Europe, the early morning hours, when overnight batch jobs have finished and the first users log in, are exactly where an extra team adds value.
Severity levels
Define severity levels before anything goes wrong. A simple four-level scheme works for most teams.
| Severity | Definition | Expected response |
|---|---|---|
| SEV 1 | Production unavailable or data integrity at risk for many users; security incident suspected | Immediate response inside the defined coverage; escalation to client incident lead at once |
| SEV 2 | Major function impaired, workaround difficult | Response within the agreed time inside coverage hours |
| SEV 3 | Minor function impaired, workaround available | Handled in the next working window |
| SEV 4 | Cosmetic or no user impact | Regular backlog |
Response times per severity should be written into the contract as service levels, distinguishing between coverage hours and on-call hours. Avoid promising response times that require heroics. A reliable 30-minute response inside agreed hours is worth more than an aspirational five-minute response nobody can sustain.
Roles during an incident
- Incident lead: coordinates the response, decides on actions such as rollback, and owns communication. For SEV 1, this role usually sits on the client side, because it involves business decisions and customer communication.
- Technical responder: diagnoses and fixes. This can be the external team, within the access rights agreed for incident work.
- Communications owner: updates internal stakeholders and, if needed, customers. Always on the client side for external communication.
- Scribe: records the timeline as it happens. Often overlooked, and essential for the postmortem.
Write down who fills each role during each coverage window, including deputies. During an incident, nobody should need to ask.
Access for incident work
Incident response requires production access, which is the most sensitive access in the system. A sensible pattern combines least privilege with speed:
- Read access to logs, metrics and traces for the responding team at all times.
- Pre-approved runbook actions, such as restarting a service, scaling, toggling a feature flag or rolling back a release, available to the responder with logging.
- Broader production access only through a break-glass procedure: time-limited, automatically logged and reviewed afterwards.
- No production personal data copied out of production for diagnosis.
Reporting duties in two jurisdictions
Incidents involving personal data trigger DSGVO duties. Under Art. 33 DSGVO, the controller must generally notify the supervisory authority within 72 hours of becoming aware of a personal data breach, unless it is unlikely to result in a risk to individuals. A processor must notify the controller without undue delay after becoming aware of a breach (Art. 33 Abs. 2 DSGVO). The contract with the external team should translate this into a concrete internal deadline and contact path, well inside the 72 hours.
A provider based in India may also have its own obligations. Under directions issued by CERT-In in April 2022, specified types of cyber incidents must be reported to CERT-In within six hours of being noticed. Clarify at contract stage how such duties of the provider interact with your own processes and confidentiality requirements, so that both sides know in advance which notifications may be required.
On-call outside coverage hours
If you need coverage beyond the combined working hours, treat on-call as a separately specified service with its own scope, response times and fees, rather than an informal expectation. Define which severities trigger a call-out, who is reachable through which channel, and what the responder may do without further approval. Rotations should be planned so that no individual carries a disproportionate share, and the provider should organise its own rota.
Requests should reach the external team through the agreed incident channel, typically an alerting tool and a defined escalation path, not through ad hoc messages from various client employees. This keeps response predictable and keeps the provider's organisation of its own staff intact, which matters for a Werkvertrag or Dienstvertrag structure.
Postmortems
Every SEV 1 and SEV 2 incident should produce a written, blameless postmortem within a few working days:
- Summary and impact: what happened, who was affected, for how long.
- Timeline: detection, response, mitigation, resolution, with timestamps in one stated time zone.
- Causes: technical causes and contributing factors, including process and communication.
- What went well, what did not.
- Actions: concrete, owned, dated, tracked in the backlog.
State all times in UTC or in one agreed local time zone. Mixed time zones in an incident timeline are a classic source of confusion in distributed teams.
Rehearse before you need it
Run a tabletop exercise at least once early in the engagement: walk through a realistic SEV 1 scenario in the overlap window, with roles, access and communication as written. Every point where someone hesitates is a gap in the model. With any provider, the essentials are the same: written roles, pre-approved actions, clear reporting paths and blameless learning.
The information on DSGVO, CERT-In and contract structures is general information, not legal advice.
An incident shows within minutes whether a distributed model was designed or merely assumed. In a niogod™ engagement, the Projektleiter maintains the incident contact matrix and the Senior Consultant owns runbooks and postmortem quality on the provider side, so roles are written down before the first alert fires. If extended coverage across time zones sits on your roadmap for next year, it is a conversation worth having early.
Talk to niogod™
Planning engineering capacity in DACH? Book a 30-minute consultation with the niogod™ team.
Book consultation