Yes. For an IT Director, I’d treat an MSP selection less like “finding a company that answers phones” and more like outsourcing a critical IT operating function. The biggest mistake is choosing on price and a polished sales presentation rather than on measurable service quality, security, and accountability.
CISA specifically recommends that organizations establish a master requirements list, define a shared-responsibility model, and put specific performance SLAs into the MSP contract.
1. Define exactly what you're outsourcing
Before talking to MSPs, establish your service boundary.
What should an IT Director look for when hiring a managed service provider to take over help desk responsibilities?
Google AI Mode· Oct 15, 2026
Hiring a managed service provider (MSP) to handle help desk responsibilities is a major shift. As an IT Director, you are outsourcing your front line of user experience, so you need a partner who protects your brand, metrics, and sanity.
Here is what you should look for during the evaluation process:
Service Level Agreements (SLAs) & Metrics : Look for granular, realistic SLAs that match your business hours (e.g., speed to answer for phone/chat, first-contact resolution rates, and ticket closure times). Ensure they offer financial penalties or credits if they miss targets consistently.
Cultural Fit & Communication : The help desk is the face of IT to the rest of the company. Their technicians must communicate clearly, exhibit empathy, and match or respect your organization’s internal culture and tone.
: Determine how they handle shift coverage, holidays, and sick days. Ask whether they use onshore, nearshore, or offshore staffing, and ensure the English proficiency and technical competence align with your users' expectations.
How can I identify a reliable managed service provider to take over our enterprise help desk functions?
Don't let the MSP define your requirements for you. A good service desk should be a single point of contact for incidents and requests, with defined SLAs and measurable performance.
2. Require hard numbers—not "great customer service"
This is probably the most important part of your RFP.
Ask each MSP to commit to metrics such as:
Responsiveness
Average speed to answer
% calls answered within X seconds
First response time
Abandonment rate
Resolution
First-contact resolution %
Mean time to resolve
Tickets resolved within SLA
Reopen rate
Escalation rate
Customer experience
CSAT
NPS, if they actually use it
User complaint rate
Executive/VIP satisfaction
Operational quality
Ticket documentation quality
Knowledge-base utilization
Automation/self-service rate
Backlog aging
Problem-management activity
Don't accept something like:
"We target a 95% SLA."
Ask:
95% of what, measured how, over what period, excluding what circumstances, and what happens if you miss it?
ITIL's current guidance explicitly emphasizes service levels, operational performance and continual improvement rather than simply operating a ticket queue.
3. Make them show you the service desk
Don't just take a sales demo.
Ask to see:
The actual ticketing portal
A real ticket from creation through closure
Escalation workflow
Knowledge base
Reporting dashboard
SLA dashboard
Major incident process
User onboarding workflow
After-hours escalation
How tickets are routed to L2/L3
How they communicate during an outage
And ask:
"Can I speak with two customers who have a similar employee count, technology stack and support model to ours?"
Ideally, speak to the references without the salesperson present.
Ask those customers:
What does the MSP do poorly?
What happens when they miss an SLA?
How good are their technicians?
How frequently do you have to escalate?
Do tickets actually get resolved, or just closed?
How painful was implementation?
What did the MSP promise during sales that didn't materialize?
Would you hire them again?
That last question is remarkably revealing.
4. Investigate the people behind the service
An MSP can have excellent technology and mediocre people.
Ask:
Where are L1/L2/L3 technicians located?
Are they employees or contractors?
What percentage is offshore?
What is technician turnover?
How long do technicians stay on your account?
Will you have a named service delivery manager?
What is the technician-to-user ratio?
How are technicians trained?
What certifications are maintained?
How do they handle holidays and sick leave?
What happens when your assigned engineer quits?
I'd particularly investigate technician turnover. A constantly changing help desk destroys institutional knowledge and makes users feel like they're starting from scratch every time they open a ticket.
5. Treat security as a separate evaluation
This is where MSP selection gets serious.
An MSP may have privileged access to:
Microsoft 365
Entra ID
Active Directory
endpoints
servers
firewalls
backups
cloud infrastructure
security tooling
CISA explicitly identifies MSPs with root/superuser/admin access as a supply-chain risk requiring appropriate vetting.
I'd require answers to:
Identity & access
MFA for all privileged access?
Privileged Access Management?
Separate admin accounts?
Just-in-time access?
Least privilege?
Customer-controlled accounts?
Full audit logging?
MSP security
SOC 2 Type II?
ISO 27001?
Penetration testing?
Security awareness training?
Background checks?
Incident-response plan?
Cyber insurance?
Your data
How is customer data segregated?
Where is it stored?
Who can access it?
How long is it retained?
What happens when the contract ends?
Can you obtain your complete ticket/history/configuration data?
CISA also recommends addressing incident responsibilities, outage compensation, data separation, logging/records, and security requirements contractually—not merely trusting the vendor's verbal assurances.
6. Don't let them force you into their technology stack
This is a common MSP trap.
Ask:
"What ITSM platform will you use, and who owns the data?"
Then determine whether the MSP wants to move you onto its preferred RMM/PSA/ITSM ecosystem.
That's not inherently bad. But you should understand the implications.
You want the ability to export:
Tickets
Users
Assets
Configuration information
Knowledge articles
Documentation
Reporting data
If switching MSPs requires rebuilding your entire IT environment because the incumbent owns the operational data, you've created unnecessary vendor lock-in.
7. Examine the commercial model carefully
Compare total cost, not the headline "per user/month."
Ask about:
Per-user pricing
Per-device pricing
Minimum monthly commitment
After-hours charges
Project rates
On-site support
Travel
Hardware handling
Microsoft licensing
Security products
Backup
New employee onboarding
Termination fees
Emergency work
Out-of-scope work
Then give every bidder the same hypothetical scenarios:
"We acquire a 75-person company. What does onboarding cost?"
"We have a ransomware incident at 2 AM. What's included?"
"We need 30 laptops imaged and shipped in two weeks."
"Our CFO can't access M365 at 7:30 AM."
"We open 400 tickets in a month instead of our normal 250."
You'll quickly discover which vendors are genuinely offering a managed service and which are selling an inexpensive ticket queue and charging extra for everything else.
8. Pay special attention to the SLA
Your contract should specify things such as:
Priority 1
Definition
Response time
Escalation time
Communication frequency
24×7 coverage
Executive notification
Major incident process
Priority 2/3/4
Response
Resolution/restore targets
Escalation
But don't make resolution time the only metric.
An MSP can game resolution statistics by closing tickets prematurely.
I'd include:
First response
Time to restore service
Resolution time
First-contact resolution
Reopen rate
SLA compliance
CSAT
Backlog
Escalation rate
And specify service credits or other remedies for persistent failures.
CISA recommends specific performance-related SLAs and clear delineation between operational IT and security responsibilities.
9. Require a transition plan
The best MSP in the world can fail if onboarding is terrible.
Make the proposal include a 30/60/90-day transition plan.
I'd expect something roughly like:
Days 0–30
Discovery
Asset inventory
Application inventory
User inventory
Documentation review
Network/security assessment
Ticket baseline
Knowledge transfer
Days 31–60
Tool deployment
Process integration
Knowledge-base construction
Runbooks
User communications
Shadow support
Days 61–90
MSP assumes operational responsibility
SLA measurement begins
Optimization
Executive reporting
Continuous improvement
Don't allow them to say:
"We'll figure out the transition after contract signature."
10. Use an objective scorecard
I'd weight an evaluation approximately like this:
Category
Weight
Service quality / SLA model
25%
Technical capability
15%
Security
15%
People & support model
15%
References / track record
10%
Transition plan
10%
Technology / automation
5%
Price / TCO
5%
Notice that price is only 5%.
That's intentional.
If you're outsourcing your help desk, saving $8/user/month isn't much of a victory if your employees spend 20 minutes fighting the service desk every time they need help.
My "red flag" list
I'd be very cautious if an MSP:
🚩 Won't provide customer references
🚩 Won't disclose where support personnel are located
🚩 Can't give technician turnover statistics
🚩 Has vague SLAs
🚩 Defines success primarily as "tickets closed"
🚩 Charges for virtually everything outside basic tickets
🚩 Can't demonstrate its reporting
🚩 Has no meaningful security documentation
🚩 Requires excessive privileged access
🚩 Won't explain subcontractors
🚩 Doesn't have a serious onboarding methodology
🚩 Has no major-incident process
🚩 Won't commit to data portability at termination
🚩 Has a long contract with weak termination rights
🚩 Has a different team selling the service than delivering it
And perhaps the biggest one:
🚩 The salesperson is dramatically more impressive than the people you're actually going to receive as technicians.
A very effective evaluation exercise
Rather than asking vendors to give you their standard presentation, I'd send all finalists the same 5–10 real-world scenarios and make them demonstrate how they'd handle them.
For example:
Scenario 1: Employee cannot access Microsoft 365. Scenario 2: CEO's laptop is compromised. Scenario 3: 150 users lose access to a critical application. Scenario 4: New employee starts Monday and HR submits the request Friday. Scenario 5: 25 laptops need replacement. Scenario 6: MSP discovers a critical vulnerability on 200 endpoints. Scenario 7: Major incident occurs at 11 PM Saturday.
Have each MSP walk your team through:
Detection → ticket → triage → escalation → communication → resolution → documentation → reporting → problem management.
That will tell you vastly more than a 60-slide sales deck.
One final recommendation
I'd also distinguish between outsourcing the help desk and outsourcing IT leadership.
As IT Director, you ideally remain accountable for:
strategy, architecture, vendor management, security governance, technology roadmap, budget and business alignment.
The MSP should make your organization more operationally efficient, not become an opaque replacement for IT leadership.
If you want, I can take this one step further and build you a complete MSP RFP/RFQ package—including the requirements, 100-point scoring matrix, SLA schedule, security questionnaire, reference-check questions, and a vendor comparison spreadsheet structure. That would give you something you could send directly to 5–8 MSPs.
Evaluating and selecting a managed service provider (MSP) for your help desk is a massive leverage point for your team—done right, it frees up your internal resources for strategic initiatives; done wrong, it becomes a daily exercise in ticket escalation headaches and user frustration.
Here is a structured framework to help you evaluate, vet, and choose the right partner:
Define the Scope and Service Level Agreements (SLAs) First: Map out your current ticket volume, peak hours, and required coverage (24 × 7 vs. business hours). Establish concrete metrics for First Response Time (FRT), Mean Time to Resolution (MTTR), and First Contact Resolution (FCR) rates.
Evaluate Security and Compliance Posture: Since the help desk handles password resets, credential verification, and often frontline triage, ensure the MSP aligns with your regulatory requirements (HIPAA, SOC 2 Type II, ISO 27001, GDPR). Ask about their internal security protocols and multi-factor authentication (MFA) enforcement.
Assess Tech Stack and Integration Capabilities: Determine how they plug into your ecosystem. Check if they integrate smoothly with your existing PSA/RMM tools or ticketing systems (like ServiceNow, Zendesk, or ConnectWise) or if you will be required to use their proprietary platform.
Scrutinize the Cultural Fit and Communication: Your users won't care if the MSP is technically proficient if the agents are rude or struggle with English/local nuances. Request sample call recordings or chat transcripts, and ask about their agent retention rates (high churn at the MSP means constant retraining for your environment).
Understand the Pricing and Contract Structure: Look closely at how they bill—whether it's per-user, per-ticket, or tiered flat-rate pricing. Watch out for hidden fees regarding onboarding, after-hours support, documentation updates, or project-based escalations.
Check References and Peer Validation: Don’t just rely on the case studies they hand you. Ask for references from current clients of a similar size and industry to yours, and check peer communities or review platforms like G2 MSP Reviews or Clutch MSP Directory for unfiltered feedback.
To help narrow this down, could you tell me:
Your approximate user count and monthly ticket volume?
Whether you need 24 × 7 support or standard business hours?
I can help you draft a tailored RFP or build a scorecard matrix for your vendor evaluations.
Staffing Models & Follow-the-Sun
Security & Compliance Standards : They will have deep access to your ecosystem. Verify their security posture, including SOC 2 Type II certification, multi-factor authentication (MFA) enforcement for their staff, background checks, and compliance with frameworks relevant to you (HIPAA, GDPR, PCI-DSS).
Ticketing & Tooling Integration : Evaluate how smoothly their ticketing system, remote management tools, and knowledge base will integrate with yours. Will you have real-time visibility into metrics, or will you be flying blind?
Scalability & Flexibility : Your business will grow or experience seasonal surges. Ensure they can scale up seat counts or expand hours quickly without degrading service quality or hitting rigid contract blocks.
Continuous Improvement & Reporting : A good MSP doesn't just close tickets; they analyze trends. Look for regular strategic review meetings where they present data on recurring root problems, user friction points, and recommendations for proactive fixes.
If you're currently evaluating vendors, let me know:
What size is your user base?
Do you need 24/7 coverage or standard business hours?
Are there any industry-specific compliance requirements you have to meet?
I can help you build a tailored RFP scorecard or a list of interview questions to ask candidates.
ChatGPT Search· Oct 15, 2026
What an IT Director Should Look for When Hiring a Managed Service Provider (MSP) for Help Desk Support
When outsourcing help desk responsibilities, an IT Director should evaluate an MSP as an extension of the internal IT organization—not simply as a vendor that answers tickets.
The right partner should deliver measurable service quality, strong security, a positive end-user experience, and reduced operational burden, while giving the internal IT team more time to focus on strategic initiatives.
I would evaluate prospective MSPs across 10 areas, with particular emphasis on service-level commitments, escalation quality, security, staffing, and accountability.
1. The 10 key evaluation criteria
1. Help desk experience and service quality
Look for a provider with a proven track record supporting organizations similar to yours in size, complexity, industry, and technology stack.
Demonstrated experience supporting your Microsoft 365, Entra ID, endpoint, VPN, business applications, and collaboration tools.
Clear processes for password resets, onboarding/offboarding, endpoint issues, software support, and recurring incidents.
References from current clients with comparable support requirements.
Documented quality assurance and customer satisfaction programs.
Ask: “Can you walk me through three recent tickets—from intake through resolution—and explain how you measured the user's experience?”
2. Service-level agreements (SLAs) and accountability
Avoid vague promises such as “fast response” or “24/7 support.” Require defined, measurable commitments.
Coverage hours, holidays, and after-hours support.
Response and restoration/resolution targets by priority.
Clear definitions of what starts and stops the SLA clock.
Escalation timelines, reporting cadence, and remedies for sustained underperformance.
Ask: “What happens when you miss an SLA repeatedly, and how is that reflected in your service reviews or fees?”
3. Security and identity management
Help desk personnel often have access to systems that can reset credentials, change permissions, or expose sensitive information. Their processes are part of your security perimeter.
Strong identity verification before password resets or MFA changes.
Least-privilege access, privileged-access controls, and auditable administrative activity.
Secure remote support and endpoint management.
Documented incident response, breach notification, and access revocation.
Relevant security assurance, such as SOC 2 reports or ISO 27001 certification, where appropriate.
Ask: “How do you prevent social engineering attacks against your own service desk, especially fraudulent MFA resets and executive impersonation?”
4. Staffing, expertise, and escalation
Determine who will actually support your employees—not just who presents during the sales process.
Defined Tier 1, Tier 2, and specialist escalation paths.
Sufficient staffing during peak demand and leave periods.
Named service delivery manager and technical escalation contacts.
Staff training, turnover metrics, and access to senior engineers.
Clear boundaries between the MSP and your internal IT team.
Ask: “What percentage of tickets are resolved at first contact, and who takes ownership when a technician cannot resolve an issue?”
5. Reporting and performance transparency
You need independent visibility into performance, not just a vendor's verbal assurance that service is going well.
Require a monthly dashboard covering:
Ticket volume, backlog, and aging.
First response and resolution times, segmented by priority.
First-contact resolution and reopen rates.
SLA attainment and exceptions.
User satisfaction (CSAT) and recurring incident trends.
Root-cause analysis and corrective actions.
Ask: “Can I see an anonymized monthly service report from an existing client, including missed SLAs and corrective actions?”
6. ITSM integration and process maturity
The MSP must work within your governance model rather than create a disconnected support operation.
Integration with your ticketing platform, CMDB, asset inventory, and monitoring tools where practical.
Agreed incident, request, problem, and change-management workflows.
Consistent categorization, prioritization, knowledge articles, and documentation.
Clear ticket ownership and history when cases move between teams.
A documented approach to recurring incidents and automation.
Ask: “Will we retain ownership and access to all ticket data, knowledge articles, workflows, and service history?”
7. End-user experience and communication
A help desk can meet response-time targets while still frustrating employees. Assess the entire support experience.
Easy access through phone, email, portal, or chat.
Clear, empathetic communication and realistic updates.
Support for remote, hybrid, and office-based staff.
Appropriate language coverage and accessibility.
Employee feedback loops and complaint escalation.
Ask: “How do you measure user satisfaction independently of ticket closure, and what happens when an employee is dissatisfied?”
8. Pricing, scope, and cost predictability
Compare the total cost of service, not just the advertised per-user rate.
Per-user, per-device, fixed-fee, or consumption-based pricing.
Included versus excluded services, such as projects, onsite visits, after-hours work, and application support.
Minimum commitments, annual increases, onboarding fees, and overage charges.
Charges for tickets, new locations, acquisitions, or changes in employee count.
Clear assumptions about what remains your internal team's responsibility.
Ask: “What are the five most common reasons a client receives charges outside the monthly fee?”
9. Transition and knowledge transfer
Outsourcing can fail during the handoff even when the provider is capable.
Discovery of your environment, support history, dependencies, and recurring issues.
Documentation of standard operating procedures and escalation contacts.
A phased transition, pilot period, and parallel validation where appropriate.
Defined acceptance criteria before the MSP assumes full responsibility.
A backout plan and continuity arrangements if the transition falters.
Ask: “What is your 30-, 60-, and 90-day transition plan, and how will you demonstrate that the service is ready before cutover?”
10. Scalability, continuous improvement, and exit strategy
Choose a provider that can adapt as your business grows without making you dependent on proprietary processes.
Ability to scale for acquisitions, new locations, seasonal peaks, and workforce changes.
Automation and self-service that reduce repetitive tickets without sacrificing support quality.
Quarterly improvement plans and evidence of reduced recurring incidents.
Contractual data ownership, export rights, documentation delivery, and transition assistance.
Clear termination rights, offboarding timelines, and revocation of all MSP access.
Ask: “If we terminate the agreement, how will you transfer our tickets, documentation, configurations, credentials, and operational knowledge to us or our next provider?”
2. Use a weighted scorecard to compare MSPs
I recommend using a weighted scorecard so the decision is based on evidence rather than the quality of the sales presentation. The weighting below is a suggested starting point, not an industry standard. It reflects the priorities of an IT Director outsourcing the help desk.
MSP vendor evaluation
(100 points)
Rate each category from 1 (poor) to 5 (excellent). Use evidence from proposals, references, security reviews, and demonstrations.
Weighted result
/ 100
[Reset scores]
Important: A high overall score must not compensate for a critical security weakness, unacceptable contract terms, or an inability to meet mandatory coverage requirements. Treat these as pass/fail gates.
3. Red flags that should concern an IT Director
Watch for these warning signs during vendor presentations, reference checks, and contract negotiations.
Red flag
Why it matters
SLAs promise only response times
A ticket can receive a quick acknowledgment but remain unresolved for days.
The MSP won't provide performance reports
You cannot independently verify service quality or identify trends.
Shared or excessive administrator privileges
A compromised MSP account can expose more of your environment than necessary.
Security claims without evidence
Marketing language does not demonstrate effective security controls.
Frequent technician turnover
Knowledge loss can increase resolution time and frustrate employees.
Low headline price with numerous exclusions
Security deserves particular scrutiny. The UK National Cyber Security Centre's guidance on managed service providers recommends reviewing access privileges, incident notification, accountability, recovery planning, and contract termination provisions. CISA likewise emphasizes MFA, least-privilege access, monitoring, and contractual security requirements.
4. Validate the vendor before signing
Do not rely on proposal documents alone. Require the shortlisted MSPs to demonstrate how they operate.
Review three customer references. Prioritize organizations of similar size and complexity. Ask whether the clients would renew, what service problems they experienced, and how the MSP handled a major incident.
Run a practical scenario exercise. Give the MSP realistic examples: an executive locked out of email, a suspected compromised account, a new employee starting tomorrow, and a business-critical application outage. Evaluate triage, verification, communication, escalation, and ownership.
Inspect a sample performance report. Look for actual response and resolution data, aging tickets, repeat incidents, customer satisfaction, and documented improvement actions.
Have Security and Procurement review the agreement. Verify the security schedule, incident notification requirements, liability allocation, subcontractors, data handling, service credits, renewal terms, and exit assistance.
Use a controlled transition. Agree on a baseline, pilot or phased handoff where feasible, acceptance criteria, and a fallback plan before the existing support arrangement ends.
5. Define success before the MSP takes over
Establish a baseline using your current help desk's historical data, then agree on improvement targets with the selected provider.
KPI
What to establish
First-response time
By ticket priority and support channel
Resolution time
By priority, issue type, and complexity
SLA attainment
Percentage meeting the agreed SLA
First-contact resolution
Percentage resolved without escalation
Reopen rate
Percentage of closed tickets reopened
Backlog and aging
Number of unresolved tickets and their age
Set numerical targets only after reviewing your environment, ticket volume, operating hours, and current performance. Define the measurement method, exclusions, reporting cadence, and accountability in the contract or service schedule.
My recommendation
For an IT Director, I would prioritize security and access controls, demonstrable service quality, and contractual accountability over the lowest price.
The strongest MSP is one that can show how it will support your employees, protect your environment, measure performance, and improve service over time. It should also leave you in control of your data, documentation, and ability to change providers.