1. Exploring workbook lineage
Now that we've covered the basics, let's walk through some examples of parent-child relationships for Oakmark's analysis.
In this workbook, we started building a solution for the Account Support team, and pretty early on, we added a table filtered to data for their team alone. We also removed certain columns from consideration that weren't important to this team, and created some additional calculations.
From here, we've built a chart and a summary table based on that filtered and formatted parent.
Without applying any filters directly, the chart and summary table only shows Account Support data. That's because the filter we applied in the parent table automatically flows into the child.
However, note how this filter is only applied to our chart. We can always layer filtering, formatting, and calculations depending on what our analysis calls for, but in general, it's a best practice to apply the filter the earliest place possible to improve performance and reduce the number of spots we'd need to update if we need to make a change.
Along those lines, it's a best practice to keep your lineage as short or narrow as possible to reuse the same parent element multiple times vs bringing in a new element each time you want to use that source.
For example, this page's lineage is not very efficient because if I need to change a column or add a filter, I need to do it in [X] places. Instead, this page's lineage is more efficient because when I need to make a change, I can do it here.
There might be scenarios where you will need to create multiple branches due to the analysis, but generally, it is best to keep your lineage as short or narrow as possible.
So you may wonder, why not have the shortest lineage possible and source my chart directly from my warehouse?
Technically, you can do this in Sigma, by updating the Data Source drop-down in an element to point to the data warehouse table, but we can't interact and edit the data in our warehouse connections the same way we can with table elements. It is better to create summary tables, charts and pivot tables as child elements of a table element you have brought into your workbook.
Take this lineage for example, notice that the icon next to CALLS_LOG indicates that it is the data source table in the cloud data warehouse. So, what if we want to add a new calculation to flag calls longer than 15 minutes? We'd have to add it in each chart instead of in a single parent element, since calculations cannot be added directly to data source tables.
That's why it's best to have a sort of pit stop table between your data source and your end result analysis. That gives our workbook a jumping-off point for other branches of work.
So far, we've seen how filters, formatting, and formulas flow from parent to child, and how strategic use of pit stop tables and short lineages can make our workbooks easier to manage.
These concepts may seem small at first, but they play a big role in how scalable and maintainable your workbooks become, especially as your use cases grow.
Up next, you'll test your understanding of these core ideas with a short quiz.
2. Let's practice!