Formulas is a tool that allows you to perform powerful calculations that can provide granular insight into user behavior.
The Formulas report provides a time-series analysis across events and their properties.
This article provides the following use cases you can study and adapt for your own use cases.
- Encouraging users to make in-app purchases
- Creating ad partnerships to drive users to accomplish business goals
- Evaluating two-sided marketplaces
Encouraging Users to Make In-App Purchases
In this use case, let’s examine the efficacy of games to encourage users to make an in-app purchase.
In the graph below, the green line shows the total number of games played against the total number of in-app purchases.
Certain games might encourage your users to make in-app purchases than other games.
Using Formulas, you can easily create another ratio that specifies the type of games played as a property of the event ‘Game Played’, and divide those games against the total number of in-app purchases.
When you bookmark and compare it against the aggregate ratio, you can see below that Game Type A (represented by the blue line) encourages more users to make in-app purchases.
With this in mind, you can alter the games that users play from the beginning to take on characteristics similar to that of Game Type A to drive in-app purchases.
Creating Ad Partnerships to Drive Users to Accomplish Business Goals
Formulas can help determine at a glance the overall success or failure of an entire project over time.
Autodesk determined that the ultimate key performance indicator (KPI) would be a ratio comparing the final step, ‘Saving/Sharing of an Image’, against the first step, ‘Home Screen Impressions’, of a conversion funnel.
By comparing the KPIs of different ad campaigns over time, they were able to not only gauge each campaign’s effectiveness at organically converting users, but also derive insights on the design of ad campaigns moving forward.
For more on how Autodesk leveraged Mixpanel to make data driven decisions, check out our office hours with them below.
Evaluating Two-Sided Marketplaces
Formulas can help you monitor crucial metrics that allow you to act swiftly in response to unpredictable market forces.
If you host a two-sided marketplace, be aware of the shifting behavior of your users. By dividing specific categories of search results against the total number of search events, you can see the shift in the types of products prospective buyers are interested in over time.
In the graph below, the search events with search results of Housewares and Furniture as properties are added together, and divided against all search events to give a better sense of user interest in Home Furnishing items.
What if you want to go a step further and see if your buyers are ultimately being sold the products they really wanted?
For example, suppose you track an “Item Posted” event with the property “Item Type.” The Item Type property represents the types of products your suppliers posted on the site. You could then divide a “Purchased” event (with the same property “Item Type”) by the "Item Posted" event to see whether the products met the demand.
In another example, suppose the ratio is below 1 for the “Item Type” : “Accessories”, you can then investigate further whether the accessories posted on your site do not fit the preferences of your buyers. And if so, you can acquire more accessory suppliers that fit the demand.