YOU ARE NOW READING
SMART SLAs For Business-IT Alignment
SLA (Service Level Agreement) has been a widely used and abused acronym in the industry. This is because of lack of understanding on what exactly it means. SLAs provide CIOs an effective tool to manage their relationship with service partners and businesses.
SLAs are the contractual commitments for meeting minimum performance criterion for delivery of service by the service provider to the customer. For example "End User Satisfaction" could be a SLA requirement and expected level could be 83 percent to be measured on a half yearly basis.
The contractual commitment is inked in the form of a separate SLA agreement, data points, measurement criterion, and expected levels, frequency of measurements, exceptions and penalty and rewards clause for not meeting or over achieving the criterion. Typically, the SLAs ensure that the services are being delivered as expected and should have the depth and diversity to cover all areas of the services.
Measuring Performance, Not Perception
The key to successful partnership is measuring and evaluating the partnership across verifiable facts and data. In the absence of data, perception takes over. SLAs provide customers and service providers the opportunity to measure and analyse service delivery and ensure that customers get the right services and CIOs can take objective decisions.
Agreement on SLAs does not often guarantee the improved services. This is a common failure which occurs due to lack of correct alignment of SLAs to the services agreement and alignment of service catalogue with business. In addition, service levels need to improve over a period of time and this should be guaranteed in the form of Service Level Improvement agreement.
SLAs should be SMART, as they should be Simple, Measureable, Actionable, Realistic and Time-bound. Since an SLA is the key to relationship between the customer and the service provider, each SLA objective should be carefully measured for months, analysed and negotiated and finally closed as an agreement. This is a common best practice across the industry.
Service providers rarely agree to upfront SLA targets at the time of signing initial deal or entering into a relationship. Service providers usually ask for measurement period, data and validation exercise before they sign SLA contracts. This makes it complicated, as you are negotiating something after you have entered into a relationship.
It is further complicated by the absence or dispute over data, as the way in which service provider and the customer looks at data could be different from SLA perspective. This requires good amount of negotiation skills, drive, commitment, maturity and the desire for a win-win from both sides to close the deal.
In my experience, the chemistry between people involved in this negotiation plays a key role. As a best practice, we should spend a good amount of time in drafting a catalogue of agreed SLA objectives (not targets) much before we enter into a relationship so that the discussion on SLAs becomes objective and stays within the defined boundaries. This provides service partner clear targets and visible goal posts, once the parties enter into a relationship.
All negotiations for SLAs should close in a time bound manner and it is good to put governance around it. At times, we should define slabs for achievements over a period of time rather than upfront achievements.
SLAs and Penalty
It is always good to define two sets of SLAs - key and critical. Key SLAs are important and do not have penalty while critical SLAs are those which are very important for business and attract penalty for performance. Customer should always retain the right of switching SLAs from key to critical.
This leads to an interesting question. Is penalty sufficient to ensure the right behavior on the part of the service provider? My experience tells us that it is not as no one wins when we fail to achieve. For each penny customer penalise service partner, actual business impact to customer's business could be several times higher. Neither Service partner wants to be penalised nor would customer like providers to fail.
To motivate the provider, the customer should also incorporate the provisions to allow the provider to come out of the penalty situation by consistently exceeding expected service levels for consecutive months or some other criterion where both recover from such situations. Stringent penalty norms and regime do more harm than good to customers and their business.
It will be worth mentioning here that CIOs & IT leaders should be very realistic in setting SLA targets or objectives in SLA. Quite often, customers expect the service providers to meet the targets which they themselves never did. It is quite clear, if you have not done it, it is unlikely your provider will be able to do it on the day you transition the services but the same can be possible over a period of time.
Another aspect is rewarding for overachievement. This works better than penalty for the customer. This motivates the service provider but customers should tune the reward to business achievements so that customers provide a percentage of business outcomes which they generated from overachievement. This requires good amount of thinking of what should be the reward structure.
Traditionally, IT departments had tendency to implement SLAs, which are less business facing and more towards IT performance. Measuring SLAs from business improvement perspective helps CIOs align IT objectives with business objectives. This is also necessary to measure business performance of technology implementation and not IT performance.
Next Generation SLAs
The next generation of SLAs will go one step further to business outcomes rather than business performance and would be linked either to top line or bottom line or both. Currently, this remains a journey for all of us as the maturity against this will evolve over a period of time.
In my humble opinion, CIOs should move beyond SLAs to business Key Performance Indicator (KPIs) benchmarking and link it to balance scorecard. One way is to look at business service management, process maturity assessment, and automation prior to moving towards business outcome measurement and KPI reporting.