IT Metrics
June 26, 2007

This Interchange focused on metrics that companies are using to assess customer service satisfaction and overall performance of IT services.

Issues:
Business Hours Calculated
Best Practices
Customer Satisfaction
Define Value
Expectations
What Metrics are used by other companies
Tracking
Outage Reporting
Presentation of Data

Business Hours Calculated:
- Depends on what service is being measured
- Define hours based on specific service, meaning some services may be 24 x 7, and others may be during business hours only. First you need to know the IT service being delivered and what maintenance windows IT wants to reserve for each service.
- Depending on request – have X amount of time to handle issue
- Calculate meaningful trend time
- Much easier to report, if simply using calendar days. May have to ‘pad’ the SLA targets of days to take into account weekends, holidays, etc.

Tracking:
- Pulling info from tickets
- Take a poll from time to time from the user/business community to see their perception of what main issues there are. Also, see what issues are being resolved by business area “super users” who end up solving a lot of issues for their peers. These “super users” should be incorporated into the Service/Help Desk as extensions of support – not as direct reports, but as advocates and associates that can help point our trends.
- Turn into business sense

Values:
- Align IT with business goals and business processes. Connect the big picture and communicate it – IT services should be measured as to the value they add to business process.
- Do customer satisfaction surveys
- Do business unit management surveys on how IT services enable their business processes and help meet their business objectives.

Customer Satisfaction:
- Don’t ignore customer. Stay in touch with them. Communicating is key. Even if you don’t resolve the issue in time, communicating constantly, honestly, and timely will keep customer satisfaction higher.
-“Important Customer” – When ticket comes in from them, everyone at help desk gets notified. This customer has a senior manager or technical service manger assigned to them.
- Have SLA (Service Level Agreement.) Be careful because customer will want everything.
- Put service level target in your SLA after negotiating with the business areas where the users are located.
- User needs to know what is in SLA. Communicate the agreed service level targets often and clearly.

Internal Measure:
- Establish business rules associated with severity levels
- Level 1, 2, 3, & 4. Each level has a time of when it should be handled
- Vary clock by what issue is
- Make sure severity is accepted by customer. (Initial severity can change during the life of the issue).
- Consider using priority system made up of Impact (how extensive is the damage that’s being done to the business) by Urgency (how fast does IT have to resolve this issue).
- Measure not only MTTR (mean time to resolve), but number of touches.

Business Objective - “Was Customer Satisfied”?
- Incident Management is about Work Arounds - Get user back in business as soon as possible.
- If necessary get issue resolved first and then go back to fixing the root of the issue. The business objective should be to restore the current business interruption, AND then find and eliminate the root cause.
- Use routing call to technicians in different tiers. Consider using skills based routing.

Address at Staff Meeting:
- Difficult in assigning blame for how long it takes to fix issue, if call passes through several levels.
- Why is call going through so many people?
-What can be done to fix this problem?
- How do you assign proper level?
- How do you hold one person accountable?

OLA - Operating Level Agreements are internal. This is an agreement between separate areas within IT to get the different areas to be accountable for helping to attain service levels. The OLAs support SLAs.

Underpinning Contracts are contractual agreements between any third party supplier of IT support to your IT Organization. Unlike the SLAs and OLAs the Underpinning contract is an actual contract.

Relationship Manager is a person that knows what IT needs and why, and also understands the business requirements. It is a part time or full time (depending on size of company) position. They need to be a good communicator and be willing to take a beating from both sides. They are the one that brokers negotiation of service level targets between IT and the business.

Rework:
How to reduce rework: Set up your filter and have automation get info you want.
Keep it simple
What is problem
Why did it happen
Steps taken to fix it

Setting up metrics takes time. You need to be sure that data source is correct and that the measurements are correct. This may take a few months to get right and get your “Baseline” of metrics to measure improvement against.

Outage Calculation:
- It is hard to measure outages. It is either up or not.
- Better to measure outage in minutes than percentages.

Cost:
- Business Impact Analysis – looks at all of the business processes / services, and determines what cost impact to the company would happen with the interruption of these processes / services. Then, those that are determined to be Vital Business Functions are looked at to find what IT services enable them, and what impact from those IT services being interrupted.
- What is cost impact to company
- Has to be site specific (HQ, Remote, Satellite)

Presentation of Data:
- Charts
- Write reports / Leverage tools
- Use Excel
- Have info wiki to start
- Don’t just report the bad – report the good things that IT is doing in business terms.

Best Practices:
- Measure only what matters
- Have to decide what matters to your specific need
- How to present meaningful metrics (Lost X amount of dollars)
- Put date behind dollars you are presenting
- You need: Mission; Goals; and Objective. Objective is time period to meet goals that are in place to satisfy your mission.

- KPI is performance indicator for objectives.
- Strategies are ways to fulfill objectives
- Goals are long term:
- What signs change metrics: Use the Balance Scorecard, Cost/Productivity, Internal customer stat, External customer stat, organizational maturity.

User of metrics need to give you what they want to see and what format, and what they will do with the metrics.

Resources:
Gartner
Forester
ITIL Books & itSMF books on metrics
Balance Scorecard
White Papers
Atlanta Chapter of Help Desk Institute

Comments/Benefits:
Changing Metrics
Outage Metrics / How it’s calculated differently
Presentation/To clearly present data
Input from IT Managers/IT Managers are my customers
What other companies are doing/Diversity of input
Best practices / Great analogy by Steve
Networking
Customer Survey
Measuring customer satisfaction. It’s my job
Reporting level
Internal vs. External
Customer satisfaction/Current project
Customer Impact