E-commerce descriptive analysis meets Business Intelligence.

Key metric 1: Measuring Revenue

  • A. Statistics on revenue by product type, which item has the highest revenue?

    B. Statistics on revenue trends by day/month/quarter/year?

    C. Find out why revenues decline at certain times of the year.

    D. Proportion of revenue by region and province?

  • Using implicit calculation for some visualization limits the flexibility and makes it harder to interpret. An example is the daily revenue, which fluctuates highly daily. Thus, I decided to process the data & calculation using DAX.

    The key function I use for moving the average for revenue, orders, and mean_price using CALCULATE, AVERAGEX, and DATESINPERIOD.

  • A. The card of comparing monthly sales allows for quick assessment on sales performance

    B. Daily Revenue & 7DMV shows the revenue trends over long performance

    C. The 7DMV of price and orders help determine the main influence on revenue trend. In this case, the significant drop in orders cause the revenue decline.

    D. The % revenue by ordered provide helps the user identify demand based on location. For example, see how regional preference/demand changes over time.

    E. The order, price, and revenue by product category help with clustering types of products in terms of business strategy (Price focus or Quantity focus).

Unfortunatly the published Power BI dashboards just doesn’t work


Other dashboards: Ensuring delivery timeliness

  • 1.1 What is the % of orders that are completed and shipped?

    1.2 What is the % of late shipped orders?

    1.3 Does province or product group affect the % above?

    2.1 What is the % of ship & receive locations in the same province with late orders? 

    2.2 Ratio % of ship & receive locations other than regions of late orders.

    3.1 Is there any relationship between order shipping distance and shipping time?

Key metric 2: Keeping track of orders & demands

  • A. Where do orders usually come from?

    B. % demand by provinces and what are the top provinces & cities with the greatest and smallest demand?

    C. Which product is ordered the most? 

    D. Product type over Province

    E. Relationship between number of orders and # undelivered packages.

  • Most of the requirements are forward and can be done with implicit measures except for most units sold in the city.

    There are two parts of the card 1 is the title name that change to include the city name and then the number of orders below.

    The key functions I used are CALCULATE, FILTER, COUNTROWS, MAXX, MINX, MAX, MIN, COUNT

  • A. The card of most units sold in (city) gives a quick sense of which city is the most important

    B. Likewise the card of the least units sold in the city with the least importance.

    C. The % Demand by province horizontal bar graphs allow us to know which province would be a good place to start a marketing campaign or to estimate the market penetration.

    D. The number of orders by each province graph can show us which region may need more funding or focus on distribution network and storage.

    E. The number of delivered and not delivered by months shows the general trends between the two. We may need to investigate the issues if there is any divergence in the general path.

    F. Lastly the number of orders by product type & province. There is a tooltip for this visual that is not shown here, showing which product is the most popular for each province, to help with distribution.

  • 1.1 Avg delivery time for orders? (group by different amount of days or lateness)

    1.2 How to distribute that index according to shipping location and product?

    2.1 What is the location of warehouse distribution and proportion across provinces?

    2.2 Top warehouse with large % of the total number of shipping orders?