ANALYTICS-CON-301 practice question 3 of 15

A client builds a dashboard that presents current and long-term stock measures. Currently, the data is at a daily level. The data presents as a bar chart that presents monthly results over current and previous years. Some measures must present as monthly averages. What should the consultant recommend to limit the data source for optimal performance?

Full set: 55 questions
Answer options
A: Limit data to current and previous years and leave data at daily level to calculate the averages in the report.
B: Limit data to current and previous years, move calculating averages to data layer, and aggregate dates to monthly level. Correct answer
C: Move calculating averages to data layer and aggregate dates to monthly level.
D: Limit data to current and previous years as well as to the last day of each month to eliminate the need to use the averages.
Correct Answer:
B. Limit data to current and previous years, move calculating averages to data layer, and aggregate dates to monthly level.
Explanation

For optimal performance, data processing should be shifted from the query layer (dashboard) to the data preparation layer (Recipe or Dataflow). Aggregating daily data to a monthly level drastically reduces the number of rows in the dataset, which is the most significant factor for query speed. Pre-calculating the required monthly averages in the data layer ensures the dashboard only has to render results, not compute them on the fly. Filtering the data to only the necessary years (current and previous) further reduces the dataset size, adhering to the principle of loading only the data required for the analysis. This combined approach minimizes dataset size and query complexity, leading to the best possible performance.

Why Incorrect

A. Leaving data at the daily level forces the dashboard's query engine to process a large volume of rows to compute averages at runtime, which is highly inefficient and slow.
C. This option is good but incomplete. It fails to filter out older, unnecessary years, resulting in a larger dataset than required and, therefore, suboptimal performance.
D. This approach is fundamentally flawed as it provides incorrect data. Using the last day of the month is a snapshot, not a monthly average, failing to meet the business requirement.

References

1. Salesforce Help - CRM Analytics Data Preparation Guide - Aggregate Transformation: "Use the Aggregate transformation to roll up data to a coarser granularity. You can aggregate data to get the sum, average, count, minimum, or maximum of a measure column... Aggregating data can significantly reduce the number of rows in your dataset, which can improve query performance." This supports moving calculations to the data layer and aggregating to the monthly level.

2. Salesforce Help - CRM Analytics Data Preparation Guide - Filter Transformation: "Use the Filter transformation to remove rows that you don’t need in your dataset... For better performance, filter data early in the recipe to reduce the number of rows processed by subsequent transformations." This supports limiting the data to the current and previous years.

3. Salesforce Help - Optimize CRM Analytics Dashboards - Work with Your Datasets: "The number one factor for dashboard performance is the number of rows your queries must process... To reduce the number of rows in your datasets, use a filter transformation in a recipe to remove unneeded data. If you don’t need granular-level data, such as at the second or minute, roll it up." This reference holistically supports the combined strategy in option B.

Scroll to Top