BLOGBUSINESS INTELLIGENCESAP ANALYTICS CLOUD

The BI & Analytics team of McCoy & Partners are experts in SAP Analytics Cloud (SAC). We would like to share a few insights on how to create a story in SAC and what you should take in consideration before starting your story creation.

First of the creation of a SAC data model which is required for the creation of a dashboard/story. A SAC model could be based on a flat file, SAP data sources like BW or Universes, SQL databases, app data like S/4 HANA or Google Big query and SAP live connections (BW or HANA). For this blog the model was created based on an excel file.

Before creating your model it’s recommended to think about what you would like to report in your dashboard as changing the model in SAC is not yet that flexible. Therefore taking a bit of time to think about your model saves you the time of re-creating your model a few times.

Things to consider when creating a model:

  • Type of charts you want to create, e.g. year on year comparison

  • Time dimension, e.g. timestamp, date or a dimension

  • Dimension names, currently names can’t be adjusted after model creation

  • Key-figures/calculations, e.g. aggregation level, number of decimals

    • General rule: have your calculations as low as possible in your model/DWH except for percentage calculations

In this example one of the visuals to be created was a year on year (YoY) comparison.

e.g.:

Figure 1: YoY comparison (v2)

To be able to create the YoY comparison you need a separate time dimension as the time dimension itself doesn’t allow a YoY comparison as you can’t select year and month as a separate dimension.

To realize the split year/month dimension we used the fiscal year/period dimension. At first the dimension was used as is and wasn’t split. In the SAC model the fiscal year period dimension was enriched with a few additional fields (Month, Year, Sort_id).

Figure 2: Fiscal year/period dimension SAC data model

These fields were used in the report to create the YoY chart, but we ran in to the problem with sorting as SAC was sorting the Month dimension alphabetically. To solve this we enriched the model with Sort_id and added this dimension to the chart:

Figure 3: chart structure YoY chart

Result:

Figure 4: YoY chart (v1)

At first this was meeting the requirement and worked quite fine but as the Sort_id couldn’t be hidden like in WebI and a custom sort option is not (yet) available in SAC the look and feel was a bit off.

In a few of Jacob Stark’s video’s about SAC/Dashboard creation you see in the intro the YoY comparison just with month names which kept us thinking there must be another way to create the YoY chart but couldn’t find a solution online.

So back to the drawing board and think about a solution which brought us the idea of splitting the fiscal year/period dimension into year and month. To realize this, you have to use the split transformation when loading data in the model to split the fiscal year/period dimension into year and month.

Figure 5: Split transformation SAC model.

Using these separate dimensions you will be able to create the YoY comparison as seen in figure 1.

Creating a story went quite fast as we are highly experienced with other front-end tools like SAP Web Intelligence, Tableau and Qlik. Like in the other tools creating charts is as easy as selecting one or more dimension(s) and one of more key-figure(s) depending on the chart type. SAC will tell you when the selected fields do not meet the criteria of a chart.

As described above is that sorting is still a bit limited, there is no custom sort ability (yet), you could work-around this in your data model.

As the release cycle of new content is every 2 weeks we are looking forward to the new developments/content which are coming!

 

Undocumented features / Header scaling

One of the challenges we faced was scaling of the chart headers when scaling down to a smaller resolution (e.g. on a lower resolution laptop or beamer). To solve this: you can view your story in device preview mode:

Figure 6: Device preview

Figure 7: display mode

This will provide you the opportunity to have an additional styling option called front scale:

Figure 8: Font scaling option

Figure 9: KPI header without scaling

When you use this function to scale the headers to fit the screen but even with this function we had some challenges. One of the challenges we faced was that the scaling didn’t change back to normal format when returning to normal view mode.

Figure 10: Font size in designer mode after scaling in device preview mode

One of the resolutions to solve this is by creating separate text fields for the headers and scaled these to the desired size.

Figure 11 & 12: KPI with loose txt header without scaling and with scaling

In case you would like to know more about SAP Analytics Cloud please do not hesitate to contact McCoy & Partners.

Thank you for reading this blog!