What is Period-over-Period analysis?
1. What is Period-over-Period analysis?
Welcome back. In this chapter we'll learn how to conduct period-over-period analysis in Sigma.2. What is period-over-period (PoP) analysis?
Period-over-period analysis (often called "PoP" or Pop analysis) is a way to compare values across timeframes. The idea is simple: you look at a value for one period of time - like this month3. What is period-over-period (PoP) analysis?
and compare it to the value for a previous, similar period such as the prior month or the same month last year, to understand increases and decreases over time.4. Defining comparison periods with Oakmark
Perhaps Oakmark Bank wants to do period-over-period analysis to understand trends in monthly transaction counts. The first challenge when conducting period-over-period analysis is defining what time periods to compare to each other.5. Choosing the best timeframe for comparison
Taking the month of January as an example, they could compare January transactions to December transactions.6. Choosing the best timeframe for comparison
or they could compare January of this year to January of the prior year. Each option has pros and cons, and the correct choice depends on the desired insight, and industry-specific seasonality.7. Comparing short-term and long-term data
In Oakmark’s case, comparing January to December would help them understand short-term, month-to-month volatility, but isn’t the best choice for understanding long-term trends, since banking activity can be volatile month-to-month, especially around year end. Instead, comparing the transaction count for January of this year to the transaction count of January in the prior year would be a better choice.8. Aligning timeframes with analysis goals
Sigma offers developers many options for comparing time periods, so consider your specific use cases carefully when selecting comparison timeframes. In all PoP analysis, the important thing is comparing apples-to-apples. We would never compare a month's worth of data to an entire quarter's worth of data. Thankfully, Sigma makes this intuitive.9. DEMO
Now let’s see a simple period-over-period comparison in Sigma. Imagine we want to compare Oakmark’s transaction volume in Q1 to Q1 of the year before. This table has 8 quarters of historical data. Let’s add a comparison column to each row, showing the number of transactions for the same quarter in the prior year. Using the `Add new column via’ option on this column’s drop-down, we see that we can access the period-over-period comparison menu, or, Sigma has a quick option to immediately add a comparison based on the last quarter or same quarter last year. The readymade option is a quick way to add the value from the comparison time period, but as we'll see, the period-over-period comparison menu is just as easy and can add a bit more detail. Let's try it. We’ll add a new column, this time using the available period-over-period comparison menu. Now we can see that for Q1 of 2025, the transaction count for Q1 of 2024, about 428K, has been added, as well as some metrics showing that this represents an increase of 16% for Q1 of 2025 compared to Q1 of 2024. We’ll cover this comparison menu in more detail in the next video. For now, you see that with only a few clicks, Sigma has added the relevant history without doing any complex table joins or lookups! Notice that at some point in the table, we run out of history, and we stop seeing comparison values. Sigma developers must consider how null rows might impact any child elements from this table, such as a line chart. A final note on period-over-period analysis in Sigma. The methods we'll cover do not rely on your data being sorted chronologically. Instead, Sigma uses intelligent date lookups, and finds the appropriate time period even when it isn't a predictable number of rows away. For example, if we sort this table in a non-chronological manner, we see that the comparison values for Q1 2025 stay the same. Similarly, if we have a table like this one, where a few months of data are missing, Sigma will still find the matching period. Before we go deeper into period-over-period analysis, test your understanding with a few exercises.10. Let's practice!
Create Your Free Account
or
By continuing, you accept our Terms of Use, our Privacy Policy and that your data is stored in the USA.