When searching for “customer success health score” you’ll find quite a few suggestions. But same as with Customer Success Playbooks most of them haven’t adopted yet, the data age.
The most advanced customer success software solutions are still suggesting that each company will identify a set of metrics that they believe matter. The examples include:
- Product usage
- NPS Survey
- Customer success manager subjective opinion
- Financials aspects
Each metric would get a fixed manual weight that will be rolled up to constitute the final customer success health score.
There are few flaws in this approach:
- Selecting the metrics is based on other companies best practices and internal consensus reached in long debates and not that much data.
- This approach does not address dynamic metric weights. For example, if product usage drops below a certain threshold it is not possible to give product usage higher weight in the overall score.
- It lacks a clear linkage to the customer success playbooks and thus cannot provide a prioritized guidance to customer success managers.
- And the most important aspect, it does not even attempt to predict the risk of churn or the opportunity to expand.
The purpose of customer success health score is to surface the likelihood to achieve organization goals (reduce churn, increase expansion) for each account.
The AI-driven approach
This approach is based on the following:
- Clear definition of goals
- Gather relevant data signals including mainly: (a) engagement of your team (Customer Success, Sales and Delivery) with your customers. (b) Product usage data – breadth and depth (c) Customer service cases.
- Based on account segmentation, score each account with likelihood to retain and/or likelihood to expand.
- Modern AI models are not “blackbox” that just provide predictions i.e. score. They should also provide an explanation that allow the team to take corrective measures.
Some of the explanations correlate directly to customer success playbook items and some are insights discovered by the model. It should be possible to easily add these explanations to the playbooks and let the team execute on them.
The debate about what is the best way to define customer success health score can be ended. AI models can provide a pretty accurate health prediction and call to action guidance.
Here is an example of Komiko’s Customer Success Health Score.
AI driven customer success health score – summary
The advantages of the AI driven approach include:
- Accurate health prediction based on your data signals
- Clear call-to-action guidance that will impact account health score
- On-going dynamic improvement of the AI model – the model keeps monitoring and learning behavior patterns including, success/failure signals, engagement, product usage, users and customers behavior.
- Full alignment with the customer success playbooks
Check out Komiko Customer Success Solution, if you’d like to further discuss, see a demo or start a free trial.