Health Scoring - Part Deux - Your input invited

  • 3 August 2016
  • 4 replies
  • 71 views

Userlevel 7
Badge +2
Hey everyone!



Health scoring is one of those things that is always being tweaked and adjusted. And just when you think you have it dialed in, it's probably time to take another look.



Why? Well, the landscape of your business and your customers is ever evolving. Contributing factors to customer risk 1 or 2 years ago are likely still impacting your customers today, but maybe not to the same degree, and new factors may now be present, or at new weightings.



Add to this the changes in data streams and validity components that are available to you. All this means that tweaking your health score from time to time is a very good thing. 



So.....I ask thee, noble community member: what are some recent efforts you've made to fine tune your health score? Please share!

4 replies

Badge
I agree, the health scoring and factors should continue to be reviewed for accuracy, relevance, and to determine if other factors should be included as part of the overall score.



After running for numerous months with a set of objective and subjective health measures, we are making a few changes.   In addition to specific health measures changes, we are driving to have more

measures and percentage of the overall health score be objective.  And

have less of the overall health score be controlled (weighted) by a subjective

measure.  



To get real specific, here's a few examples of some changes we are making after some re-evalution of our overall health scoring.



1.   For our support measure, instead of looking at sheer number of cases for different types of customers.   We are now coming up with a support measure for each customer based on several factors:   lack of cases entered, service level target of case responses, average age of cases, case survey responses, number of high priority cases entered over a time period.   After evaluating our customer data and churn data, this support measure is a better way for us to view our support world versus just a number of cases count.



2.   We created a new product management measure based on product bugs and enhancements entered by each customer.   This particular measure and trend is used in our risk and overall health discussions with the Product Management team.



3.  We created a new "engagement" measure which accounts for items such as, technical contact product certifications, training, webinar/event attendance, community participation, and advocacy.   We have taken some of the subjective measure weighting and moved that weighting to this new objective measure.



hope that helps provide a couple points and specific examples of health scoring 'tweaks".  Im glad to further discuss the details and how we went about rolling out the changes.



twix
Hi Tom, 



Thanks for sharing! Could you explain how you incorporate so many factors into one score, 



Thanks,

Angela
Badge
we did some work in SFDC;  some custom formula fields and custom code.



For example, Support Measure...



- we have 5 sub-factors we calculate in SFDC

- we start the overall Support Measure at 100

- then reduce the total by a certain amount based on the 5 factors

- the subfactors are things like SLAs, CSAT, TTR on cases.

- lets say the Avg CSAT for one account is low, then that subfactor would reduce the Support measure by 20pts.

- lets say the Avg TTR for one account was xxx, then that subfactor would reduce the Support measure by 10 pts.

- lets say there were NO SLAs missed for one account, then NO points are taken away from the Support Measure.

- after all the subfactors are taken into account, then the Support Measure is 100 or less depending on the subfactors.



A bit complex to explain, but Im glad to further explain if you ping me or want to have a call.
That makes sense Tom, thanks so much for taking the time to explain it!

Reply