Tags: FILTER

Customer analysis by Country and time period

{2 Comments}

Here is a Sales dataset of 8 columns and 29 rows.  It basically details the revenue earned and cash collected by service type, Customer, Country and Period.  For a selected Country and time period, there could be customers availing of both services or of any 1 service.


There are 2 broad questions that one may want to get answers to:

  1. Determine the number of customers who availed of a certain number of services
  2. Determine customers with whom business was forged for the first time and those who churned out

For a chosen country and Year/Month, the first question stated above further sub-divides into:

  1. How may customers availed of both services - Consultancy and Implementation
  2. How may customers availed of only one of the two services

So if a user selects the Country as India and Year/Month as January 2015, then Customers who availed of both services would be 1,3 and 4.  Note that Customer 2 should not be considered (even though he/she availed of both services) because the revenue earned from one of the services (Implementation) was nil.  For the same selection (India and January 2015), the Customers who availed of only 1 service would be Customer 2 - this customer availed of only the Consultancy service (Revenue was earned from this Customer only for this service).  After applying a filter on the source dataset, the rows for India and January 2015 are:

The expected result is shown below in PowerBI desktop software.  If you are not concerned with who those customers are (you just want the count), then you may simply remove the Customer Name field from the visual.

The second question is to determine the number of new and lost customers.  If a customer was not in the database in any prior month, the customer is identified as new.  To clarify, a customer who availed of the Consultancy service in a prior month also availed of the Implementation service for the first time in the current month would not be counted as a new customer.  If a customer ceases to generate revenue in any month, the customer would be counted as lost (churned) in that month.  So when USA is selected in the Country slicer and Year/Month is February 2015, the expected result is:

I have solved this question with the help of the PowerPivot.  You may download my PowerBI desktop solution file from here and source Excel workbook from here.  This problem can also be solved in MS Excel using the PowerPivot.

Compute Relative Size Factor per vendor

{0 Comments}

Relative size factor (RSF) is a test to identify anomalies where the largest amount for subsets in a given key is outside the norm for those subsets. This test compares the top two amounts for each subset and calculates the RSF for each. In order to identify potential fraudulent activities in invoice payment data, one utilizes the largest and the second-largest amounts to calculate a ratio based on purchases that are grouped by vendors.  You may read more on this topic here.

Here is a 3 column dataset.  The first column is Vendor Number, the second is Invoice number and last is invoice amount.  There can be multiple invoices per vendor.  The objective is to determine the highest invoice value for a vendor and divide that by the second highest invoice value for that same vendor to get a ratio.  The same needs to be done for all vendors.  An interesting case in the dataset below is Vendor_No V4439 - there are 2 instances of highest value for this vendor i.e. 25,378.30 and another 2 instances of second highest value i.e. 24,068.25.  The RSF for this case will be 25,378.30/24,068.25.  If there is no instance of second highest value for a vendor, then the result should be 0.

The expected result is:

I have solved this question with the help of the PowerPivot.  You may download my solution workbook from here.

Determine the lowest bidding vendor(s) for each product in a Pivot Table

{0 Comments}

Imagine a dataset like this.  This dataset shows vendors that submitted proposals for supplying various parts to a Company.  There is one column for each of the twelve months.

untitled

Via a simple Pivot Table, one can determine the lowest bidding vendor per product (part) for any chosen month.  However, one may also want to know the names of those vendors for each product (as seen in column G below).  Notice, that Vendor 2 and Vendor 3 submitted the lowest bid for Product 1 and therefore both names should appear in the result.

untitled

I have solved this problem using PowerPivot and Power Query a.k.a. Data > Get & Transform in Excel 2016.  You may download my solution workbook from here.

Filtering on 2 date fields within one Table

{0 Comments}

This table contains a list of all the inspections created and completed within different time periods.

untitled

The objective is to create two Tables from this single table - one showing the Accounts created within the chosen time period and another showing the those that were closed within the same time period.  Here are screenshots of the expected results.

untitled1 untitled2 untitled3 untitled4

You may download my PowerBI desktop solution workbook from here.  The same solution can be obtained in Excel as well (using Power Query and PowerPivot).

Compute transaction fee based on a tiered pricing model

{0 Comments}

Consider a simple dataset as shown below:
untitled
For each tier, the tier rate is incrementally applied to the volume within the tier volume range.  Given the following transaction volumes, one may want to compute the transaction fee
untitled1
The expected result is shown below
untitled2

As one can observe, for a transaction value of 400,000, the fee has been computed as 3% on the first 100,000 and 2.5% on the next 300,000. You may download my solution workbook from here. In the file, I have shared 2 solutions - a conventional formula based one and a PowerPivot solution.

I have also solved a similar question here.

Perform an aggregation on Top x items after satisfying certain conditions

{0 Comments}

Visualise a 5 column dataset as show below.  This is a very small sample of the actual dataset.  It shows the date on which supplies were received for each item from Vendors and whether those supplies had errors in them.  Finally those identified errors have been bucketed into relevant categories.  The Item ID# is a code assigned by the Client Organisation.  The format of the ID# is X [Last two digits of year of receipt] [Four digit Number].  For items received from a certain vendor, the last four digits of any item will always be in ascending order of the Date received.  In other words. for items received from a certain vendor, the last 4 digits of an item received on 13 January 2017 will always be greater than the last 4 digits of an item received on 11 January 2017 from the same vendor.  There will never be repetitions in the Item ID# column.

Vendor Name Item ID# Date Received Error with Item? Error Category
California X170016 1-16-2017 No
California X170014 1-13-2017 Yes Labeling Error
California X170015 1-13-2017 Yes Packaging Error
California X170008 1-9-2017 Yes Quality Issue
California X170003 1-2-2017 No
California X160645 12-26-2016 Yes Packaging Error
California X160646 12-26-2016 No
California X160644 12-25-2016 Yes Labeling Error
California X160638 12-20-2016 Yes Quality Issue
California X160633 12-15-2016 No
California X160626 12-8-2016 No
California X160625 12-7-2016 Yes Packaging Error
California X160624 12-5-2016 Yes Labeling Error
California X160618 11-23-2016 Yes Quality Issue
California X160613 11-13-2016 No
California X160606 10-30-2016 No

The objective is to compute the error rate by vendor and Error category for the 10 most recent transaction dates with that specific vendor.  So, for vendor Name "California" and  Error category as "Packing Error", this ratio should be computed as = Number of packing Errors on 10 most recent dates/10.

Here is a simple snapshot of the Data for California.  I have filtered the dataset where Vendor Name is California and then sorted the Date received column in descending order.  Please note that when i filter the dataset on California, a lot more rows are returned.  I am only showing the Top 10 rows here because that is what is important for solving this question.

Vendor Name Item ID# Date Received Error with Item? Error Category
California X170016 1-16-2017 No
California X170014 1-13-2017 Yes Labeling Error
California X170015 1-13-2017 Yes Packaging Error
California X170008 1-9-2017 Yes Quality Issue
California X170003 1-2-2017 No
California X160645 12-26-2016 Yes Packaging Error
California X160646 12-26-2016 No
California X160644 12-25-2016 Yes Labeling Error
California X160638 12-20-2016 Yes Quality Issue
California X160633 12-15-2016 No

The expected result is:

Row Labels Labeling Error Packaging Error Quality Issue Factory Error
Boise 30.00%
California 20.00% 20.00% 20.00%
Detroit 70.00% 30.00%
Ekalaka 20.00% 20.00%

I have solved this problem using the PowerPivot. You may download my solution workbook from here.

Compute the average of values against the 5 most recent dates of each Category

{0 Comments}

Here is a simple 3 column dataset showing Categories, Date and Value

Catagorie Date Value
Fish 08-12-2015 6
Crab 05-12-2015 7
Crab 04-12-2015 6
Bird 27-11-2015 4
Snow 25-11-2015 10
Cat 21-11-2015 7
Dog 12-11-2015 5
Dog 28-10-2015 5
Fish 12-10-2015 3
Bird 11-10-2015 9
Dog 22-09-2015 9
Crab 17-08-2015 8
Cat 11-08-2015 1
Fish 20-07-2015 5
Crab 03-07-2015 2
Fish 02-06-2015 8
Fish 26-05-2015 9
Dog 14-05-2015 4
Snow 07-05-2015 7
Bird 03-05-2015 9
Cat 20-04-2015 10
Cat 15-04-2015 2
Snow 13-04-2015 3
Crab 29-03-2015 5
Crab 23-03-2015 2
Bird 14-03-2015 5
Cat 14-03-2015 1
Dog 26-02-2015 9
Fish 16-02-2015 4
Fish 08-02-2015 6
Bird 18-01-2015 1
Snow 06-01-2015 10

The objective is to compute category wise average of values against the 5 most recent dates.  So the expected result is:

Categorie Category wise average values on 5 most recent dates
Bird 5.60
Cat 4.20
Crab 5.60
Dog 6.40
Fish 6.20
Snow 7.50

So for the Category of Fish (dates are sorted in descending order), the average should be 6.2 (average of 6,3,5,8,9)

Catagorie Date Value
Fish 08-12-2015 6
Fish 12-10-2015 3
Fish 20-07-2015 5
Fish 02-06-2015 8
Fish 26-05-2015 9
Fish 16-02-2015 4
Fish 08-02-2015 6

I have solved this problem using the PowerPivot. You may download my solution workbook from here.

Determine the total number of projects by Status

{0 Comments}

Here's a simple 3 column table showing Date, Project name (Cat.) and Status of the project.  Each project can have multiple status entries on different dates.  So as you can observe, project "alpha_9383993" was In Progress on Oct 2, 2017, remained so on October 5, 2017 but was completed on October 6, 2017.

Date Cat. Status
02-Oct-17 alpha_9383993 In Progress
03-Oct-17 Pulse_9387388 In Progress
04-Oct-17 Pulse_9387388 Rework
05-Oct-17 alpha_9383993 In Progress
06-Oct-17 alpha_9383993 Completed
07-Oct-17 Pulse_9387388 Completed
08-Oct-17 Oppo_tes_9383 In Progress
09-Oct-17 Oppo_Max_8977 Rework

The objective is to determine the count of projects by Status as per the most recent status of every project.  So the expected result is:

Row Labels measure 2
Completed 2
In Progress 1
Rework 1

The result for In Progress should be one because there is only one such project - Oppo_tes_9383.  Project alpha_9383993 should not be counted because it was completed on October 6, 2017.  Likewise the result for Rework should be one because there is only one such project - Oppo_Max_8977.  Project Pulse_9387388 should not be counted because it was completed on October 7,2017.

I have solved this problem with the PowerPivot.  You may download my solution workbook from here.

In a Pivot Table, compute highest revenue earned on any day from each customer and the date thereof

{0 Comments}

Here's a simple dataset showing the Date of sale, Customer Name and Sales amount.

Date Customer Name Sales amount
12-03-2017 A 1
12-03-2017 A 2
12-03-2017 A 3
12-03-2017 B 4
12-03-2017 B 5
12-03-2017 B 6
12-03-2017 B 7
12-03-2017 B 8
13-03-2017 A 1
13-03-2017 A 1

The objective is to determine Customer wise:

  1. Highest revenue; and
  2. Date on which that highest revenue was earned

The expected result is

Row Labels Highest revenue Date of highest revenue
A 6 12-Mar-2017
B 30 12-Mar-2017
Grand Total 36

So, for A, revenue earned on March 12, 2017 is 6 and on March 13, 2017 is 2.  Since the higher of the two is 6, that is the result which should appear in the Pivot Table.

I have solved this problem with the PowerPivot.  You may download my solution workbook from here.

In a Pivot Table, show the most frequently appearing text entry by a certain parameter

{0 Comments}

Here's a simple two column dataset

Comment Identifier Intervals
A 3pm-6pm
A 9pm-12pm
S 3pm-6pm
S 3pm-6pm
S 9pm-12pm
A 9pm-12pm
S 9pm-12pm
D 3pm-6pm
A 9pm-12pm
A 9pm-12pm
A 9pm-12pm
A 3pm-6pm
A 3pm-6pm

For identifiers listed in column A, there are time intervals in column B. Note that for a certain identifier, a time interval can appear multiple times. The objective is two-fold:

  1. For each identifier, show the time interval which appears most frequently; and
  2. For each identifier, compute the count of the time interval which appears most frequently

The expected result is:

untitled

As you can observe, the result for S is two time periods - 3pm-6pm and 9pm-12pm.  This is because each of them appears twice.

You may download my solution from here.