The following sections also relate to data. Please consider the holistic data requirement when designing the data model:
When a data point is created in a template, this is not automatically available for reporting. Reporting data lives in a separate set of normalised tables and is populated through an ‘Extract, Transform, Load’ (ETL) process that runs when a user completes a Contract Template interview. Out-of-the-box, CLMA has an ETL that covers the UCM data model only. As such, any custom data points require a custom ETL to be created, (if these data points need to be made available to the reporting engine).
Please note that Ad-hoc reporting is catered for using ‘Search’. This is covered in Data for Search.
- Have you got any existing reporting requirements for either downstream processes? Or related to regulatory reporting?
- What are your current reporting pain points?
- Do you have a reporting tool you already use?
There are two types of reports:
- Reporting on the current state (i.e. snapshot)
- Reporting on the historic data (i.e. analytics)
The first is used by managers to help manage the day-to-day activities and the latter is used for analysing past performance in order to adapt the future processes. Depending on the stakeholder providing the requirements, you are likely to get requirements for only the one or the other. For the former, this type of reporting needs to be delivered via ‘Search’. Refer to Data for Search for more information.
For analytical reporting, the main item is to identify any data points that are not already covered by the UCM. Secondly, it needs to be understood whether this data point relates to the main contract or one of its transactions (to determine at which level the data will need to be retrieved for the report).
Once a list of custom data points has been identified, the next task is to work with a Coupa Technical Architect to help design the database / table structure that will need to be created in the CLMA reporting database. Once this has been designed, the last element is the creation of the custom ETL to populate these tables/fields. This is covered in more detail in Extract Transform Load (ETL)
With the data now available, custom reports can be developed. Please refer to Bespoke Reports for more information on this section.
- As custom ETLs are required to populate custom data points into the reporting table, it is key to flag this with Customers. The creation of the ETL is not a process they can own, so once designed, any changes to it will require a services engagement (i.e. cost) to perform. As such, it is worth making sure that all (relevant) data points have been identified. If there is a risk that further data points will be required in future Contract Templates, it may be beneficial to postpone the ETL work to the end of the project / a later phase.
- The data that can be ETL’ed pertains to contract data collected during the Interview. Data related to workflows cannot currently be ETL’ed into the reporting tables.
Things to avoid
Do not try and finalise the data model until after the Data Capture templates have been authored. The specific data points/names, etc. often change during the authoring/testing process. Once the data capture templates have stabilised, then the biggest risk of changes to the data model is over.