Operational Excellence - The Practice of Cloud System Administration - page 414

Level 3 assessment is designed to be good enough for most services. Going above that
level should be reserved for high-priority services such as revenue-generating services or
services that are particularly demanding.
When setting organizational goals, focus on fixing a problem, rather than achieving a
particular assessment level. That is, never set a goal of improving the assessment of an op-
erational responsibility for the sake of improving the assessment. That is putting the cart
before the horse. Instead, identify a problem, engineer a solution, and measure the success
or failure by whether the assessment improves. This is a subtle but important difference.
Setting the goal of an improved assessment drives the wrong behavior. It encourages
people to create spot fixes that do not solve the larger problem or to fix unimportant-but-
easy issues just to get a better assessment. It would be like paying teachers based on the
grades their students receive; such a system would simply lead to all students receiving
A+'s as exams become easier and easier.
Forthisreasonitisequally wrongtosetagoalofraising theassessment level ofall ser-
vices to a certain level. An engineer's time is scarce. A push to have a high assessment for
every organizational responsibility results in expending those scarce resources on low-pri-
ority services in the name of raising an average. Such a pursuit creates bureaucratic hand-
cuffs that strangle an organization. It is something you should never do; indeed, it is a trap
that you should only hope your competition falls into.
This point cannot be overstated. If your management ever sets a corporate, division, or
organizational goal of achieving a certain assessment level on all organizational responsib-
ilities of all services, you are to show them this section of the topic and tell them that such
a plan is a recipe for disaster.
Likewise, one should never make raises and bonuses contingent on the results of an as-
sessment. This discourages people from joining the projects that need the most help. In-
stead,encourageyourbestpeopletojoinprojectswheretheycandothemostgood.Engin-
eers are highly motivated by the opportunity to do good work. The best reward for an im-
proved assessment is not money, but the opportunity to work on the most interesting pro-
ject,ortoreceivenotorietybysharingtheirknowledgewithotherteams.Rewardsbasedon
rising assessment levels are also not recommended because often it is a major achievement
to simply retain a particular level in the face of challenging times.
20.7 Getting Started
Introducing this assessment system to a team can be a challenge. Most teams are not used
toworkinginsuchadata-drivenassessmentenvironment.Itshouldbeintroducedandused
as a self-improvement tool—a way to help teams aspire to do their very best.