Data Driven Customer Success: Accounts health and prioritization

Alexey Smolyanyy
2 min readOct 26, 2022

--

Part 1: Why Data Driven Customer Success?
Part 2: What data do we need?

Obviously, after collecting all data, we need to simplify its presentation as much as possible. These are the fundamental questions of Customer Success:

  • What is the health of a certain account or accounts segment?
  • Will a given account churn, shrink or grow?
  • What is the predicted amount of churn and shrink?

They should be able to be answered instantly, and in self-service mode, and it doesn’t matter who asks: CSM, VP, CCO or CRO.

Luckily, there’s one universal answer to all of these questions: Account Health Score.

The bad news: although the answer is universal, there’s no universal formula that is perfectly applicable to everyone. However, we can describe the general logic.

I love the concept of a “score”, i.e. a virtual number which has no physical unit and is normalized, so you can always instantly assess the position of a given customer on a standard scale and compare any two accounts with each other. For example, a scale of 1..10 or 0..100 — whatever you like. I personally prefer the latter, because it can instantly be presented as a “renewal probability” in percentages.

This is my proposal for the health score calculation:

account_health = usage_score*w1 + engagement_score*w2 + csm_sentiment*w3 + nps_score*w4

As you can see the main score (account_health) itself consists of a number of different scores. These component scores are weighted, allowing the main score to be fine-tuned for a particular use case (using wN parameters). Each score may also be calculated separately. For example, the engagement score is a weighted sum of all types of engagement activities, such as emails, tickets, meetings, etc.

Usage score is a combination of the factors explained in the previous chapter.

It’s a basic formula and we can add more parameters there, like the current account lifecycle phase, number and stage of open opportunities, etc.

The next score I recommend using is the priority score. Unlike the health score, it should be unique, i.e. you can’t have two accounts with the same priority; one is always above or below the other. Why? This gives you the ability to instantly choose the correct account if you have congestion at work.

A priority score also can be calculated in multiple different ways.

For example, by a combination of the current ARR, open opportunities ARR, health score (i.e. risk of churn), and growth potential.

--

--