Why should you define KPIs towards your Outsourced services
Any service comprehends some key requirements that must be met within certain compliance intervals.
Making it clear, here are some examples and what could be the inherent KPI:
- A call center service needs to operate during regular working hours – KPI, work time from 07:00 a.m. through 07:00 p.m.
- Facebook, Google, YouTube, LinkedIn, other … need to always be up and running – KPI, 99,99% availability
- An after-sales support team needs to clearly and assertively help the client overcoming any given difficulty 9 out of each 10 calls – KPI, 90% client satisfaction
What is a KPI then?
A Key Performance Indicator is an assertively measurable data point which accurately represents up to which extent a given service or tool is performing towards meeting inherent established objectives.
The relevancy of defining and having in place appropriate KPIs pertains the accuracy degree with which you can assess how successful a given service is.
Is the service being delivered as promised and required, worst? Better than expected?
How can KPIs be defined?
In order to define a given KPI, you must begin by setting the goal of the service component that will have such KPI associated with. And while defining a KPI, you may be compelled to create a new accessory one that complements the one you have just defined.
Out of the example mentioned above let’s pick the “Big Ones”, Google, Facebook, … availability KPI.
So, our target market (in the case all of us that use these tools), demands the service to always be up and running, so one KPI that will be part of the contract we establish with the Outsourcing provider who will ensure our IT Landscape is “service availability” and the threshold (as demanded by the target market) is 100% (always).
What is the goal?
- To have the tool available each and every time anyone accesses it, 24 hours a day, 365 days a year.
Which degree of deviation is acceptable?
- The acceptable deviation is one that is perceived by the user not as the tool being inaccessible, but the system is slow to respond. This should not happen more than once every 50 times the same user accesses the tool.
How to measure it?
- A 3rd party independent distributed monitoring system needs to be in place to constantly ping the tool. Afterward, such monitoring system registry will be cross checked with both user complaints and the service provider servers and Telekom infrastructure logs.
When should the measuring be active?
- This measuring needs to always be active with worldwide coverage.
Now we know that 100% service availability is just not possible and forcing impossible terms into a contract just bears the inevitable outcome of having both parties operating over an error, nothing else.
Nevertheless, as IT experts, we know that there are ways to assure something like 99,99% IT Landscape uptime by means of active multi mode clustering over several geographies. This is to say if the System architecture is one that operates with multiple active replicas, each with the capacity to take over the service in case the operational node goes down and being those physically installed in distinct geographies, hence mitigating potential power or communication outages; it is possible to have in place a distributed IT Landscape that can sustain such a demanding KPI threshold.
Now let’s think of which additional support KPIs would make sense to set in place beside the “availability KPI”.
It was assessed (as previously mentioned) that users will not be upset if the system seems slow to respond once every 50 times they assess it and that means 2% of the times. Let’s consider (academically) that current internet standard is a server response time of 100 milliseconds, the support KPI here can be defined as 98% of server response within 100 milliseconds of a request by the user.
Common relevant KPIs
The ahead listed IT related KPIs are relevant for the majority of businesses (when applicable) and therefore you should consider employing them in the contracts you establish with your Outsourcing provider:
- Service availability – Established the required service uptime within a given time interval.
- Average response time – The end user is a key element in any IT service, for he/ she is the client “persona”. Humans usually only call when in need of assistance and someone in need of assistance is not prone towards waiting. It is relevant to define what is the expectation towards getting an answer from the service and establishing a KPI that mirrors such market expectation.
- Change driven Incidents rate – Change is always present and something, in fact, needed under a context where continuous improvement is promoted. Nevertheless, it is most relevant that a change does not generate an incident which may compromise the service you render and client satisfaction, so it is relevant to measure how many changes lead to incidents.
- Security compliance rate – In our days, Security is one major pillar of any IT Service so it is vital to understand up to which extent one’s IT Landscape or processes are safe or represent any risk potential.
- Adherence to budgeted expenses – When you Outsource something you set a budget in place, yet some variations may happen. It is nevertheless relevant to measure how often your provider either stays below or goes above budget because your business model also depends on it.
Either the above mentioned or other that are more relevant for your business, KPIs should be considered and set in place by you towards your Outsourcing services providers. That helps both parties for it allows a set of assertive metrics that make clear if the service is being delivered as committed and mainly where there is room/ need to improve.
In fact, the main advantage of having KPIs in place does not relate to the possibility of applying penalties to suppliers that do not meet the established criteria but to understand where the service is below the intended level and define actions that can support necessary improvement.