Network and application teams that fear they fall short of industry benchmarks for performance , resiliency , stability , usage or other metrics will be familiar with a common corrective action : drawing a line in the sand , establishing ‘ what good looks like ’ and then working to get there .
Whether this definition of ‘ good ’ is enough to raise a team or organization ’ s output up to where it needs to be to properly succeed in today ’ s economic environment is a hot topic .
The business market is hyper-competitive . Market leaders haven ’ t gotten where they are by striving to meet basic measures of performance or success . They ’ re there because they set the standard for others to live up to .
In that way , being ‘ good ’ does not imply being ‘ bestin-class ’. That ’ s been the case for over 100 years : in scenario planning models like ‘ good , better , best ’, the ‘ good ’ option is the least ambitious of the three . Defining what ‘ good ’ looks like might be a first step to improvement , but it ’ s a starting point , not an end state .
That ’ s the challenge for any team or organization that is trying to set themselves up on a trajectory of improvement and growth . It ’ s not enough to be ‘ good ’ anymore . There ’ s a need to be better than ‘ good ’. Only by continuously getting better , can organizations and teams become the ‘ best ’ in their respective fields .
A new standard needs to be set that can evolve with changing business and economic conditions and that standard needs to be underpinned by endto-end visibility .
Reaching for ‘ good ’
Defining what ‘ good ’ looks like is an important first step on the journey , but teams and organizations are likely to encounter common obstacles that prevent even ‘ good ’ from being achieved .
Specifically , organizations and teams are susceptible to missing the mark on ‘ good ’ if they lean too heavily on external benchmarks or SLAs as accepted definitions of what ‘ good ’ looks like , or if they define ‘ good ’ themselves without having data to support their decision-making . It ’ s worth unpacking both in a bit more detail .
First , there are common industry benchmarks around what is considered to be ‘ good ’ – page loads in under two seconds , latency or ping times within a certain number of milliseconds , no greater than 1 % packet loss on a circuit . The list goes on . Some of these were set many years ago , have not evolved since and have lost currency . They ’ re also ‘ someone else ’ s ’ benchmarks and won ’ t fit every set of circumstances .
SLAs – Service Level Agreements – pose much the same problem . They are often set very low to make breaching them difficult and would not match up with an organization ’ s own internal definition of what constituted ‘ good ’ availability or service . For that reason , their utility as a model for setting a ‘ good ’ internal baseline is fairly limited .
Second , where organizations try to set their own baseline instead of using someone else ’ s , but where they do not have data and visibility to support the internal discussion and decision-making , they risk setting ‘ good ’ too low .
Definitions of ‘ good ’ can end up becoming watered down by what is internally perceived as realistic and achievable , within a set of known constraints – financial , time , cultural and so on . While realism is an important input into any discussion , there has to be awareness of how it can work against innovation .
Solutions Analyst Mike Hicks , Principal Solutions Architect at Cisco ThousandEyes
www . intelligentcio . com INTELLIGENTCIO APAC 37