Best Practice to Deal with BLANK in Power BI

Measure definitions where BLANKs are converted into some default or alternate values are generally considered as poor design because they are not efficient. When measures are evaluated, Power BI attempts to retrieve all the possible groupings within the current filter context and then accordingly visuals are updated. Keeping BLANK is better because the visuals by default eliminate the groupings where summarizations are BLANK and hence renders the reports faster.

If you convert your BLANK values to some default or alternate values within the measure definition then your dataset becomes large in terms of what all needs to be rendered to display the visual e.g. if you had a dataset with 100 rows, out of which say 30 rows were evaluating the measure as BLANK so in this case only 70 rows were being rendered to display the visual. But, if you convert BLANK to say zero then all your hundred rows have values and hence displayed which means time to render 100 rows is of course larger than the time to render 70 rows.

However, if there is a need NOT to display BLANK in the visual then handle that at the visual level rather than at the measure level.

So, it is recommended that your measures should return BLANK when a meaningful value can not be returned by it without any consideration to any default or any other alternate value. In a general principle too, if you hardcode a default value at the place of BLANK values then your calculation framework violates the principle of being a generic implementation.

e.g.

Better design practice with no consideration of default/alternate values

Poor design where BLANKs are converted to a default of zero