Networks

Managed Network and Security Services SLA Guide: The Eight Numbers Plus One Clause Indian CIOs Should Negotiate in 2026

Updated: May 26, 2026

Magnifying glass over coins representing SLA pricing and cost analysis
5 Minutes Read

A defensible managed services SLA names eight numbers and one clause: uptime, time to acknowledge, time to respond, time to restore, mean time to detect, mean time to contain, reporting cadence, service-credit formula, and a named Service Delivery Manager with a written escalation matrix. Restoration is the number that pays. 

For context, Gartner estimates India's managed security services segment will grow 15.1 per cent in 2026, the fastest-growing subsegment within India's security services category. This blog is the procurement-side reference; for the commercial side, see our Managed Services service page

What Does a Managed Services SLA Actually Promise? 

A Service Level Agreement is the part of the contract where the provider commits to measurable performance. Everything outside the SLA is best-effort. Most Indian buyers read the headline 99.9 per cent uptime and stop. They miss that 99.9 per cent uptime equals 8 hours 46 minutes of downtime a year, while 99.99 per cent equals 52 minutes. 

Network and security SLAs are different beasts. A network SLA centres on availability and restoration. A security SLA centres on detection, containment and investigation. A single managed services contract usually covers both. 

The Eight Numbers Plus One Clause to Negotiate 

The Eight Numbers Plus One Clause to Negotiate
Metric  What It Means  Defensible Enterprise Benchmark 
Uptime / Availability  Percentage of time the service is operational  99.9% standard, 99.99% for critical sites 
Time to Acknowledge  Time from ticket logged to provider responding  15 minutes for P1, 30 minutes for P2 
Time to Respond  Time from acknowledgement to active work  30 minutes for P1, 2 hours for P2 
Time to Restore (MTTR, Mean Time to Restore)  Time from incident to service restoration  4 hours for P1, 8 hours for P2 
Mean Time to Detect (MTTD)  Average time from event to detection in the SOC  Target under 30 minutes; under 15 minutes is a stretch SLA 
Mean Time to Contain (MTTC)  Average time from detection to containment  Target under 4 hours for confirmed incidents; under 1 hour is a stretch SLA 
Reporting Cadence  Frequency and depth of operational reports  Weekly ops, monthly exec, quarterly board 
Service Credit Formula  Penalty applied when targets are missed  5 to 20% of monthly fee, banded by severity 
Named SDM and Escalation Matrix  A single accountable Service Delivery Manager with a written L1 to L2 to L3 to SDM to Provider VP escalation ladder  Named in the contract, with 24x7 reachability for P1 

Severity definitions for the table.  

P1: total service outage or confirmed security incident affecting all users or critical systems.

P2: partial outage or degradation affecting a site, a team or a service.  

P3: single-user impact.  

P4: informational or minor request. 

Time to Restore is the number Indian buyers underweight most.  

Time to Respond is what the provider commits to in writing.  

Time to Restore is what the business actually experiences.  

Negotiate the second harder than the first. 

How Does the Service Credit Formula Work? 

Service credits convert a missed SLA into a refund on the monthly fee. Each band of miss earns a percentage credit, banded by severity and capped per month. Most enterprise contracts cap a single-month credit at 20 to 30 per cent of fees and provide a contract-exit right after three consecutive months of breach or six months in twelve. 

Worked example in INR. A managed services contract priced at ?2 crore per year (?16.7 lakh per month) with a 99.99 per cent uptime SLA. A drop to 99.95 per cent triggers a banded credit. On a four-band, 5 per cent per band structure, the credit is roughly 20 per cent of monthly fees, around ?3.3 lakh, applied automatically against the next invoice. Two clauses to demand: credits issued automatically without the customer raising a claim, and the right to terminate without penalty after a defined pattern of breach. 

What Is In Scope and What Is Out? 

The scope clause is where most disputes happen. Pin down five things in writing. The estate (every device, site, VLAN, firewall, cloud tenant and SaaS app, with serial numbers and IPs). Operating hours (24x7 or business hours). Change management (who can make changes, how many per month, and what counts as a project). Carve-outs (scheduled maintenance, force majeure, third-party outages and ISP downtime are usually excluded; most enterprise contracts hold the maintenance window between 1 and 4 hours per month). Patching (who patches what, with what notice, and the SLA on emergency zero-day patching). 

What India-Specific Clauses Must the SLA Carry? 

Four clauses that go beyond the standard MSP template. 

CERT-In reporting workflow. The provider must commit to detecting, classifying and supporting CERT-In notification of in-scope incidents within the six-hour window in the 2022 direction. 

DPDP breach support. Once Phase 3 of the DPDP Rules takes effect on 13 May 2027, the provider must support breach reporting to the Data Protection Board without delay, with a detailed report within 72 hours. 

Data residency. Logs, recordings, telemetry and configuration backups for Indian customers should be stored in India unless a documented exception exists. Specify the data centre regions in the contract. 

INR billing and tax treatment. GST treatment, withholding and indexation should sit in the commercials, not be left for later. 

What Does a Real Reporting Cadence Look Like? 

Three layers. Weekly operational reports for the SOC and NOC managers with ticket volumes, MTTR and any SLA misses. Monthly executive reports for the CIO and CISO with trend lines and a top-five risks view. Quarterly board reports that translate operational metrics into business risk. 

What About Exit and Offboarding? 

The exit clause is the most commonly weakened part of an Indian managed services contract. Demand four things. A documented offboarding playbook. A defined transition window of 60 to 90 days with full provider co-operation. Return of all configurations, runbooks, logs and credentials in machine-readable formats. A mutual no-poach clause. 

How Proactive Data Systems Helps 

Proactive Data Systems is a Preferred Partner under the Cisco 360 Partner Program across Networking, Security, Collaboration, Cloud & AI and Services. Our NOC and SOC handle managed network and security workloads for BFSI, manufacturing and ITeS enterprises, with the eight metrics, the India clauses and the credit and exit terms enterprise CIOs ask for.  

Book an SLA Audit. Ninety minutes. We benchmark your current managed services contract against the eight numbers, the named SDM clause and the India-specific clauses, and hand you a redline-ready document within ten working days.

Frequently Asked Questions

99.9 per cent for standard estates, 99.99 per cent for critical sites. 99.9 per cent equals about 8 hours 46 minutes of downtime a year. 99.99 per cent equals 52 minutes.
Response time is when the provider begins active work on a ticket. Restoration time is when the service is back. Restoration is the number that matches the business impact.
5 to 20 per cent of the monthly fee banded by severity and repeat misses, capped at 20 to 30 per cent in any one month, applied automatically against the next invoice, with exit rights after three consecutive months of breach.
A total service outage or a confirmed security incident affecting all users or critical systems. P2 is partial outage or degradation; P3 is single-user impact; P4 is informational.
They must be. The provider should commit to detection, classification and reporting support inside CERT-In's six-hour window and DPDP's without delay, plus a 72-hour cycle.
Most enterprise contracts hold it between 1 and 4 hours per month, scheduled in low-traffic windows, with 14 days' notice for non-emergency changes.
60 to 90 days with full provider co-operation, machine-readable handover of configurations, runbooks and logs, and a mutual no-poach clause.

Whitepapers

E-Books

Contact Us

We value the opportunity to interact with you, Please feel free to get in touch with us.

 

 

 

 

Share a few details to get started.

We'll get back to you shortly.