2 Replies Latest reply on May 11, 2016 4:42 AM by Sten Pedersen

    Counting of text values (availability verdicts) from a calculated formula

    Sten Pedersen

      Hi folks,

      I have case there I have a number of physical locations that provide IT services. All locations does not provide all services, some is introduced or removed over time. We measure availability of these services on each location. The availability measure (%) is provided as an average per day. A threshold indicates acceptable or non-acceptable availability (verdict) . The verdict part is calculated in Tableau and not provided. The threshold and verdict part is solved on different time period aggregations.


      Data provided is: service, location, date and availability%. Global threshold or per service will be available in separate data source (in my attached example a global threshold is used)


      One KPI defined is defined as availability on acceptable level (above threshold) in the service and location dimension. The KPI reporting is done in both the Service dimension and the Location dimension, i.e. how many services out of total that is on acceptable level for each location and how many locations that are on acceptable level for each service. E.g. if % services per location on acceptable level is above 90% = stretch level, 70% - 90% commitment level, 50%-70% Robust and <50% off-track (this should of course be configurable). Could also be number of services/locations instead of %, e.g. all services available on acceptable level = stretch, 1 one service not available to acceptable level = Commitment etc.


      I can’t find a way to count the number of services on acceptable level (i.e. verdict=above in my example). Someone that can help out in this area? Either in my solution or with other solution.


      Another KPI one could imaging is on single service and location, number of days or weeks on acceptable level per month (without visualizing the daily/weekly data/verdict). Haven’t got the time to look into that yet.


      Thanks for your help