The aim of any business is to generate revenue and, eventually, profit. Every metric has its place, but we simply don't have the time or resources to pay close attention to all of them. Sometimes metrics even seem to contradict each other. What metrics are treated as priority depends on the company's strategy and plans.
A company's or department's strategy determines which metrics need to be improved. An analyst must know how to identify the strategy and use it to choose important metrics that can be improved.
Knowing the key business metrics is crucial for any analyst. This makes it possible not only to handle exciting and complex tasks, but also to understand their importance for the company.
When changes in an indicator have a direct impact on the company's or department's main business targets, it's called a leading metric. Examples could include CAC, LTV, and conversion rate. The important thing is to make sure the metric you've chosen is actually leading. (Leading metrics help you make predictions; lagging metrics help explain the past.)
The indicator, the changes of which have a direct impact on the company's or department's main metric, is called leading. It may be CAC, LTV, conversion or something else. The important thing is to make sure you've chosen the metric which is in fact leading. In practice, you can't say for sure that other metrics will remain the same, as they're all interconnected. The ways they're connected need to be described. In other words, we have to decompose the metric.
Let's see how metric decomposition works from the following example.
Take revenue.
On the one hand, revenue = average purchase size x number of purchases
where number of purchases = number of acquired users x user-order conversion rate.
On the other hand, revenue = number of acquired users x revenue per visitor (RPV).
At the same time, revenue per visitor = total revenue / number of visitors.
The average revenue per acquired user is called revenue per visitor (RPV).
One and the same metric (say, revenue) can be decomposed in various ways, and all of them will be correct. This makes it possible to study the funnel from different angles to find ways to improve the key business metric.