Success Metrics are custom KPIs measured for each of your accounts. They are powered by equations you define, which can take into account almost any of the data you pump into Vitally, including product events, account and user traits, total users at an account, and more.
While Vitally gives you many helpful metrics out of the box, Success Metrics help you to define KPIs specific to your product and business. They are great for measuring:
Number of product events created by an account over a rolling timeframe (e.g. number of User Invited events over the last 7 days)
Percentage of an account's users performing specific event(s) over a rolling timeframe (e.g. percentage of users that have performed the User Invited event over the last 30 days)
The percentage an account has used of some feature/component they are billed for (e.g. percentage of pre-purchased licenses used)
Changes in a critical account trait over time (e.g. track how the trait
apiCallsMade changes over time)
To create a new Success Metric, navigate to Custom Account Metrics -> Success Metrics in your account settings (or get there via Quick Jump).
To create your Success Metric, you'll need to define the following:
Name: The name of your Success Metric
Equation: This is where you'll define the equation behind your Success Metric. All Success Metrics must have a numerator. Adding a denominator is optional.
Timeframe: Most Success Metrics require a rolling timeframe to measure the metric over (e.g. the number of events performed by users over the past 30 days).
Events are the individual interactions users have with your product (as tracked by your analytics service) - e.g.
Purchase made. You'll usually have at least one Success Metric that uses events since they are the primary way to signal that a specific feature has been used by a user.
Because events happen hundreds to thousands of times a day, Success Metrics that use events are measured over a rolling timeframe (up to the last 30 days). This way, your event-based metrics can focus on the customer's recent usage of your product and how that trends over time. Keep in mind that Vitally already tracks all-time event totals for all your events by default.
You can also leverage specific details about each customer's users in your Success Metrics. This is commonly helpful if your analytics service already tracks some basic stats about your users - e.g. each user has a
Total purchases made trait that tracks the total number of purchases ever made by a user in your product. In that case, you can use those stats to calculate metrics like the total and/or average number of purchases ever made for all users at each customer.
Finally, you can also leverage account traits in your Success Metrics. There are two common use cases for doing so:
1. To normalize your Success Metric based on the customer's subscription with your business. For example, pretend that your product allowed each customer to upload up to 5, 10, or 20 videos every month based on their purchased plan. If your analytics service attached that video limit to customers as traits, you could easily measure how many of those 'pre-purchased' uploads each customer actually uses every month.
2. To track changes in an account trait over time. For example, say your engineers have already attached a
Total Videos Uploaded trait to your customers. You could simply define a Success Metric that tracks the value of that trait as it changes. This way, you can easily see how that value trends over time for each account in Vitally.
Every Success Metric you create adds a new column to your accounts that you can view and report on (like all other account columns).
One unique advantage to Success Metrics: the daily value of each account's Success Metric is tracked and cached in Vitally over time. This allows you to dynamically filter accounts (and run automated workflows) by a percentage increase or decrease in a Success Metric value over a dynamic timeframe.
You can also view this historical cache of Success Metric values for a specific account in the Trends tab of their dashboard.