Normalization

Normalization shows the percent of change between comparison reports, which is useful when date ranges have a different number of days, or different volumes of traffic. Normalization takes all metrics and forces them to equal proportions, raising or lowering individual line items according to their normalized total. Normalizing lets you match trends when one date is much higher or lower than the other.

For example, if one month has three more days than another, the three-day difference might cause a significant discrepancy in a monthly A/B comparison. When you normalize the data, Analytics forces the totals of each report match, and increases or decreases the values of one column to adjust for the different number of days. Normalization is available in reports with date comparisons, or the Key Metrics reports.

How Normalization is Calculated

Normalization is calculated by:

1. Comparing report totals and calculating the proportion of data.

Pages September Page Views October Page Views

Page A

350

400

Page B

200

375

Page C

25

75

Total

575

850

Data normalization takes the totals of the two metrics (575 and 850) and determines their ratio:

575 / 850 = .676

2. Multiplying each line item by the report total proportion (using the previous report, with normalizing enabled):

Pages September Page Views October Page Views

Page A

350

237

Page B

200

135

Page C

25

17

Total

575

575

Each line item in October was multiplied by .676 (as shown above). The table now reflects the same approximate amount of September's data, allowing you to compare the two date ranges more effectively.

Data normalization applies differently if the compare dates were in opposite order, such as 850 / 575 = 1.48. Meaning, September's data would inflate itself proportionately to match closer to October's data.

See Normalizing Report Data in Help.