SLAs with Penalties Don't Work! Try This Model Instead.

Here's an excerpt from my book, "The Vendor Management Office: Unleashing the Power of Strategic Sourcing."  I've been on the sales side of the business (the "dark side") and I know for a fact that SLA penalties (OK, you have to call them liquidated damages to make judges happy, but let's face it--they're penalities) are very often baked into the P&L of the deal.  In other words, vendors tyically don't care when they have to pay a penalty for a missed SLA.  I've found a solution, I've implemented it numerous times, and it works like a charm.  Check it out.  Sorry, it's not formatted as nicely as in the book...

Service Level Agreements
While Value for Money or “VFM” metrics attempt to measure the more intangible aspects of a vendor relationship, customers will clearly be focused on what they can measure—tangible metrics in the form of service level objectives (SLOs) or service level agreements (SLAs).  Both SLOs and SLAs define a customer service expectation as well as a service standard for the vendor.  The major difference between the two is that an SLO is merely an objective whereas an SLA typically includes the payment of penalties by the vendor for not achieving service levels (the “stick”).  Less frequently, an SLA includes the payment of rewards by the customer when the vendor exceeds service levels (the “carrot”).  However, the intended effectiveness of penalties for missing service levels or rewards for exceeding service levels is questionable, and there may be a more effective alternative available to drive vendors to achieve service levels.

The Stick Model
There are two primary reasons why customers request service level penalties: to recover business costs associated with a missed service level or to induce the vendor to achieve service levels through negative reinforcement via the payment of penalties in the form of liquidated damages.

Generally, business costs resulting from missed service levels are difficult to calculate and, consequently, are not a common motivation behind requests for service levels with penalties.  On the other hand, customers frequently request service levels with penalties because they believe that penalties induce vendors to achieve service levels in order to avoid payment of the penalty.  Additionally, customers may believe that they are due a financial “discount,” in the form of a penalty, if the customer is not receiving the level of service for which they contracted.  In some cases, customers request increased penalties for sustained periods of recurring non-performance.  These increased penalties are usually capped, weakening customer leverage if vendor non-performance is systemic.

In reality, the stick model of service levels with penalties has limited effect.  In fact, it may have a negative effect on the customer in terms of pricing.  Even where vendors are generally capable of achieving a service level with an associated penalty, vendors commonly increase prices to mitigate any possible, although unlikely, penalties.  This incremental increase in price converts into an accrual on the vendor’s books for the potential expense of a penalty.  At this point, the performance-oriented behavior that the customer was attempting to drive has been diminished—whatever “hammer” or stick the customer thought they had (and probably paid for) as leverage is now a simple accounting transaction.  Further, customers must consider the possibility of an unethical vendor committing to a service level where the profit margin is still attractive because the inevitable penalties are incorporated into the vendor’s cost structure as a cost of doing business.

The Carrot Model
The carrot model assumes that financial rewards will induce the vendor to achieve higher levels of customer satisfaction.  In this model, the vendor receives rewards from the customer for consistently sustaining or exceeding service levels.  Customers can combine the carrot and stick models to incorporate a theoretical balancing factor—penalties for under performance and rewards for over performance.

This service level model has not achieved prominence as a method of driving vendor behavior due to customer-based issues such as the uncertainty of future payments (in the form of rewards) that may be committed to the vendor.  For example, if a customer did not accurately define a particular service level in contract negotiations, the customer may have unwittingly committed itself to the continuous payment of rewards for a service level that the vendor will routinely exceed.

Wrath of the Customer Model
An alternative model that can drive the desired effect is the “wrath of the customer,” or, worded in a more formal construct, service levels with escalations.  Rather than a simple accounting transaction hidden in a vendor’s financial system, escalation paths ensure visibility of non-performance to increasing levels of customer and vendor management.  Thus, each level of a vendor must answer to the next level of management when the vendor did not achieve a service level.  The effects on a vendor’s human and company behavior in this type of service level model can have a significant and far-reaching impact, with resultant tangible benefits to the customer.

The escalation model also forces “partnership-in-practice” behavior.  As much as customers hope to believe that their vendors are business partners, a vendor is, and will always be, a vendor.  Service levels with penalties tend to be contractually oriented, customer-vendor procurement relationships, where the impact of an unachieved service level is in the form of a clinical accounting transaction.  As a result, the stick model does nothing to bring a customer and vendor closer together in terms of the business relationship.  On the other hand, service levels with organizational peer-to-peer escalations tend to drive deeper relationships, with the vendor being more inclined to understand the adverse business implications to the customer of unachieved service levels.

Where the escalation path addresses a specific unachieved service level target, there needs to be a contractual mechanism, similar to an increased penalty, for addressing sustained non-performance.  In this case, the most effective remedy is a solid early termination and transition clause.  The stick in this case is the vendor’s loss of an ongoing revenue stream, rather than an increased penalty that the vendor already accounted for in its profit and loss statement.

Achieving Service Level Targets
The primary objective of the service level with escalations model is to ensure the vendor understands that their relationship with the customer centers on achieving required service levels.  The construct of this model should focus on and encompass:
  • Detailed peer-to-peer escalation procedures, to include escalation “triggers” and timeframes for escalation to the next level where resolution is not achieved
  • A pre-determined set of actions to occur when performance measurement points below minimum service level targets (the escalation trigger)
  • Clear identification of escalation levels (peers, by name and title), responsibilities, and accountability
  • The methodology for a joint problem resolution process
  • Post-mortem procedures defined for root cause analyses
  • Goals for long-term improvement
  • Acting on the intent of the contract, rather than solely the exact content
  • An operating principle of fairness, not exploitation of any contract inefficiencies
As a part of a close customer-vendor relationship, the service level with escalations model over the traditional penalties model is an alternative worthy of evaluation.  The escalations model enhances a customer’s leverage without the additional cost impact resulting from a vendor building the risk of a penalty payment into the deal profit and loss calculation.  Although a customer and Vendor Management Office staff must make a time investment and play an active role in the escalations model, this model may ultimately prove to be an effective mechanism to drive vendor performance.

 del.icio.us  Stumbleupon  Technorati  Digg 

 

What did you think of this article?




Trackbacks
  • No trackbacks exist for this entry.
Comments

  • 5/27/2008 7:33 AM Bo wrote:
    What is the service levels with escalations in details? Can you give an example, thanks.
    Reply to this
    1. 5/29/2008 12:28 PM Stephen Guth wrote:
      Bo,

      It's actually quite easy.  You would use a traditional service level objective, like "99% of the time in the monthly measurement period, vendor shall respond to all service requests within 2-business hours of the request and shall provide a resolution within 4-business hours of the request."  Just add to that the remedy, like "Where vendor fails to resolve a service request with 4-business hours, vendor shall escalate the service request to vendor's operations manager, and shall then escalate the service request to vendor's vice president of operations if the service request remains unresolved for greater than 8-business hours.  Should vendor fail to meet the service level objective for the monthly measurement period, vendor's vice president of operations and customer's vice president of vendor management shall meeting in person to discuss the root cause and vendor's action plan to improve the service level for the following monthly measurement period."  This is an off-the-top-of-my-head example, and needs a lot more work, but I think you get the picture of how it works from the example.

      Thx!
      Stephen

      Reply to this
  • 1/15/2009 6:18 PM PPC Services wrote:
    I thin "Achieving Service Level Method" with penalties would be a great model.
    Reply to this
Leave a comment

 Enter the above security code (required)

 Name

 Email (will not be published)

 Website

Your comment is 0 characters limited to 3000 characters.