Revenue Analysis

Overview


What is Revenue Analysis?

Revenue Analysis helps you maximize your profits by allowing you to trace the lifetime value of your customers and see how much they’re spending within your application. You can analyze and compare users by install date, along with the ad network, campaign and creative that brought them in. Revenue Analysis gives you the truest measure of a customer's value to your business.

How do I access Revenue Analysis?

Revenue Analysis is located in the Reports tab in the Upsight Analytics dashboard. Follow the steps below to access it:

  1. Click on "Reports" in the menu bar at the top of the dashboard.
  2. Select "Acquisition" from the navigation table on the left, then click "Revenue Analysis" from the submenu.

Revenue Analysis Button

Interface


What are the requirements for using Revenue Analysis?

To make the most out of Revenue Analysis, you should have a good understanding of the graphs used in Revenue Analysis and of terminology (such as cohorts and spending curves). The Spending Curve Plots page will give you a thorough understanding of how such plots are generated.

Upsight is proud to provide Revenue Analysis to all customers as part of the Upsight Analytics package.

Select Install Cohort

  1. Click the field under Install Date Range and a dialog will appear.

  2. Select the date range of interest (any number of days/weeks/months).

    Optional You can also repeat this query any number of times.

  3. Click Apply.

    revenue labeled screenshot

  4. Optional You can also select Campaigns, Categories, Content, and add Filters.

    revenue analysis default screenshot

    Remember to click Apply after selecting your filters.

    revenue analysis add filter screenshot

  5. Click Run to view the spending curves! They’ll show up below your workspace.

    run cohort screenshot

Workspace

You have the options of saving and revisiting your current work at a later time!

Moreover, you can also clear the workspace and export the data in the workspace into an Excel spreadsheet (.csv).

workspace screenshot

Introduction to Spending Curve Plot

The true potential of Revenue Analysis lies in the comparisons between the Spending Curve plot for All Users, and the Spending Curve plot for Paying Users. The individual graphs carry a wealth of knowledge but together, they tell a story of your business.

A. When you clicked Run in step 5, you should have generated spending curves, which are the individual lines/curves.

B. Hovering over a campaign in the workspace will highlight its corresponding curve in the plots.

C. The dots that run along the curves are called nodes; day-to-day revenue information is displayed when you hover over a node.

D. You can change the Since Install Date Range for data that is immediately presented in the Workspace.

E. The number of paying users can be visualized at the bottom of the Revenue Curve (Paying Users) graph.

filter male screenshotmale in workspace screenshot

Spending per User

  1. Calculate the Cumulative Spending and User Count for every date in range. Cumulative spending: total spending since the install date

    User Count: Total number of users since the install date

  2. Divide Cumulative Spending by User Count to find the Spending per User.

Spending per Paying User

  1. Calculate the Cumulative Spending and Cumulative Spenders for every date in range. Cumulative spenders: Total paying users since the install date

  2. Divide Cumulative Spending by Cumulative Spenders to find the Spending per User. The above is a very high-level explanation of how spending curve plots are generated.

We highly recommend that you read the next section for a complete explanation to get the most out of Revenue Analysis.

Read Spending Curve Plots to understand how spending curve plots are generated!

Understanding Results


This section provides a thorough explanation of how spending curve plots are generated.

You will get the most out of this section by reading from the very beginning to the end.

The 5-Day Spending Curve

Let’s examine the ten users who installed on June 1, 2012. We want to learn their spending behaviour during the first six days since the install.

Metric Date (2012)Total SpendingFirst Time Spenders
June 1$102
June 2$101
June 3$50
June 4$51
June 5$00
June 6$101

Using same values above, we determine the values: Days since Installation, and Cumulative Spending, and Cumulative Spenders.

Metric Date (2012)Days Since InstallationTotal SpendingCumulative SpendingFirst Time Spenders
June 10$10$102
June 21$10$201
June 32$5$250
June 43$5$301
June 54$0$300
June 65$10$401

Days Since Installation

The value for Days since Installation can vary; for this example, we have chosen a date range of 5 days for this example. However, please remember that the date of installation is Day 0—not Day 1 because a day has not yet passed since installation.

Cumulative Spending

Cumulative Spending is the total value of money spent by users since Day 0.

For instance, on Day 2 (i.e. June 3), the Cumulative Spending value is the sum of the Total Spending values of Day 0, Day 1, and Day 2.

Cumulative Spending (Day 2) = Total Spending (Day 0) + Total Spending (Day 1) + Total Spending (Day 2)

or likewise,

Cumulative Spending (Day 2) = Cumulative Spending (Day 1) + Total Spending (Day 2)

Cumulative Spenders

Similarly to Cumulative Spending, Cumulative Spenders is the total number of paying users since Day 0.

Take a look at Days 3 and 4 (i.e., June 4 and 5): the value of cumulative spenders remained at 4 because there were no first time spenders on Day 4.

Cumulative Spenders (Day 4) = Cumulative Spenders (Day 3) + First Time Spenders (Day 4)

Spending Curve Per User

Upon completion of the chart, we must calculate the Spending Curve per User before we plot this information into a graph.

The table below is slightly different from the previous ones because we are now calculating Spending per User. We already know the values for Cumulative Spending and that the total number of users, 10, has not changed.

Days since InstallationCumulative SpendingUsersSpending per User
0$1010$1
1$2010$2
2$2510$2.5
3$3010$3
4$3010$3
5$4010$4

To find Spending per User, divide Cumulative Spending by the number of Users in the corresponding row.

We can now plot the graph with Days since Installation on the x-axis (horizontal plane) and the Spending per User on the y-axis (vertical plane).

1

Spending Curve Per Paying User

Calculating the Spending Curve per Paying User is very similar to Spending Curve per User. Instead of dividing Cumulative Spending by total Users, we divide Cumulative Spending by Cumulative Spenders (paying users).

Days since InstallationCumulative SpendingCumulative SpendersSpending per Paying User
0$102$5
1$203$6.67
2$253$8.33
3$304$7.5
4$304$7.5
5$405$8

At this point, you should understand how Revenue Analysis calculates Spending Curves per User and Spending Curves per Paying User.

2

Combining Multiple Install Days

Let’s take a look at another application where we have 15 users who installed on June 2, 2012.

Again, we want to look at spending behaviour during the first six days after installing the app.

Metric Date (2012)Total SpendingFirst Time Spenders
June 2$104
June 3$200
June 4$151
June 500
June 6$51
June 7$51

Like before, we compute Cumulative Spending and Cumulative Spenders for their corresponding days.

Metric Date (2012)Days Since InstallationTotal SpendingCumulative SpendingFirst Time SpendersCumulative Spenders
June 20$10$1044
June 31$20$3004
June 42$15$4515
June 530$4505
June 64$5$5016
June 75$5$5517

Spending Curve per User

Let’s combine our Install Days from the first case (installed on June 1, 2012 group) and this case (installed on June 2, 2012 group).

As a result, our table for the combined group includes the columns Days since Installation, Cumulative Spending, Users, and Spending per User.

Days since InstallationCumulative SpendingUsersSpending per User
0$10 + $10 = $2010 + 15 = 25$0.8
1$20 + $30 = $5025$2
2$25 + $45 = $7025$2.8
3$30 + $45 = $7525$3
4$30 + $50 = $8025$3.2
5$40 + $55 = $9525$3.8

Under Cumulative Spending, each cell displays an equation that represents how we’ve combined values from two distinct install days. For example, cumulative spending for Day 2 shows $25 (Cumulative Spending for June 1, 2012 group) and $45 (cumulative spending for June 2, 2012 group). A very similar process is performed for the Users column.

After combining the two groups, we calculate Spending per User by dividing Cumulative Spending by the User Count.

Like before, we will now plot the information using Days since Installation on the x-axis and Spending per User on the y-axis:

3

By now, you should have a good understanding of how to calculate Spending per User.

We have calculated the Spending per Paying User values for you below:

Days since InstallationCumulative SpendingCumulative SpendersSpending per User
0$206$3.33
1$507$7.14
2$708$8.75
3$759$8.33
4$8010$8
5$9512$7.92

As a refresher, we calculate Spending per Paying User by dividing Cumulative Spending by Cumulative Spenders.

This is the plot for Spending Curves per Paying User.

4

Here's a refresher:

Each install day group generates a Cumulative Spending column and User counts/Cumulative Spenders (paying users) column. When the days-since-installation extend past the day in which the query is generated, some groups may have missing data as it is in the future. When combining groups, the Cumulative Spending and the User Counts/Cumulative Spenders are aggregated first, and then normalized.

Phew! Now you're ready to get cracking!