# Representative transaction KPIs - what method to chose (average, median...) / how?

In my dataset i have about 30k of transactional data. For analysis, I want to show the end user some important KPIs:

• Average of transactions with loyalty cards
• Average turnover of these transactions

The problem is, with a simple average i get "1,43..." as the result. The median of the dataset is 4. The highest number is 230, "normal" values are 10-20 transactions, most of them are 1-5.

Now i look for a solution for this problem. What would be the best to show the user a "representative" value for the average loyalty card transactions?

The Average is a good measure when a dataset contains values that are relatively evenly spread with exceptionally NO high or low values.

You can use 'Median' if you find extreme values in your dataset. It ignores the extreme values (outliers).

You may want to plot a histogram / box plot to check for outliers and decide which one to use.