Customer Health without Analytics and Automation?

Ryan Creamore

So you don't have product analytics, but you want to develop a customer health score?

It's possible, but how and when?

For a low-touch lower ACV customer base, perhaps this methodology isn't the way to go, but in this instance you're aiming to automate as much engagement as you possibly can anyway.

But, for larger organizations serving the enterprise or mid-market that haven't yet rolled out tools like Gainsight PX, Pendo, or any other app monitoring solution, a manual health score is worth the effort.

Your solution has health score metrics built into it. For example a CRM can be scored against new contacts created, new opportunities created, user adoption.

These are true outcome driving success metrics within your application. If you have the ability to login as your customer, your CSM can utilize a manual health scorecard and review the structure, adoption and utilization of the solution and score it.

Using this score, the CSM can develop initiatives to support the customer in driving the business outcomes forward. Additional training perhaps, assistance in a reconfiguration, or general business consultation where they share what other customers are doing.

How are others identifying cx health without automated tools?

Benedict Fritz

With some experience I think gut feel is worth noting and tracking. There are a lot of subtle signals that reflect how things are going that can be difficult to quantify, and don't rely on analytics or automation at all. How excited was the customer when you last talked to them? How quickly are they replying to your emails? Is their feedback generic and uninteresting, or are their feedback and thoughts ones that only someone deeply engaged with your product could have?

I've seen CSMs tracking a manual health number like this in spreadsheets, but when you ask them about it there's no single concrete factor they point to that determines it. Seems very worthwhile even if it's hard to explain!

Ed Powers

Hi @Ryan Creamore:

There's a middle step between "gut feel" and a full-blown, automated, predictive analytics solution: conduct a statistical study using regression analysis.

While objective measures are preferred, subjective CSM ratings can be used in the mix, provided there's reasonable consistency in how the ratings are applied. For a ratio (or real number) dependent variable, (for example ARR, NRR, Gross Dollar Churn), you would use multiple regression. For a nominal (categorical) dependent variable (for example, logo churn, a binomial) you would use logistic regression. These techniques yield linear equations and a reduced set of factors from which you can construct reasonably predictive Customer Health Scores. And you can do it all in Excel. 

For more on the subject and to see an example, read my blog, Using Regression Analysis to Define Customer Health Scores. If you want to learn how to do it, see my Udemy course, Make Data-Driven Decisions in Customer Success. I'm also happy to talk through these approaches with anyone interested in learning more: ed@se-partners.com 

Ryan Creamore

Ed, 

I always look forward to your posts, they are always so thoughtful and intelligent.  

Best, 
Ryan

Ryan Creamore

Benedict, 

Great feedback.  I think even in an automated health score there should be a CSM override function. 

Best, 
Ryan

Steve Bernstein

Hi Ryan -- Why not just ask the customer to assess their experience(s) and outcomes with the company, along with an overall sentiment score (such as the B2B version of the NPS "recommend" question)?  Seems to me that it's much more accurate than having the CSM guess at it, plus you can automate this process, get multiple persona involved, and get true representation from the perspective of the account. What do you think?

I *think* that the reason this isn't done more often is because companies are used to chucking surveys over the wall and suffer from poor response rates. But when treated respectfully you can achieve 80%+ response rates (lot's written in this forum and elsewhere on that but I'm happy to provide more details here as well). I just don't understand why more companies don't do this and appreciate your thoughts.
/Steve

Steve Bernstein

Hi Ryan -- Can you rely on CSMs to report it accurately? I mean, who knows better, the customer or the CSM, and could there ever be circumstances (such as performance reviews!) where the CSM is incented to "sugar coat" what's going on?

------------------------------
Steve Bernstein
Head of Voice-Of-Customer Programs at Waypoint Research Group
------------------------------
-------------------------------------------
Original Message:
Sent: 08-21-2020 15:57
From: Ryan Creamore
Subject: Customer Health without Analytics and Automation?

Benedict, 

Great feedback.  I think even in an automated health score there should be a CSM override function. 

Best, 
Ryan

------------------------------
Ryan Creamore
Director of Customer Success, North America
------------------------------

Original Message:
Sent: 08-21-2020 12:58
From: Benedict Fritz
Subject: Customer Health without Analytics and Automation?

With some experience I think gut feel is worth noting and tracking. There are a lot of subtle signals that reflect how things are going that can be difficult to quantify, and don't rely on analytics or automation at all. How excited was the customer when you last talked to them? How quickly are they replying to your emails? Is their feedback generic and uninteresting, or are their feedback and thoughts ones that only someone deeply engaged with your product could have?

I've seen CSMs tracking a manual health number like this in spreadsheets, but when you ask them about it there's no single concrete factor they point to that determines it. Seems very worthwhile even if it's hard to explain!

------------------------------
Benedict Fritz
Building the high-touch onboarding tool at https://arrows.to

Original Message:
Sent: 08-21-2020 12:27
From: Ryan Creamore
Subject: Customer Health without Analytics and Automation?

So you don't have product analytics, but you want to develop a customer health score?

It's possible, but how and when?

For a low-touch lower ACV customer base, perhaps this methodology isn't the way to go, but in this instance you're aiming to automate as much engagement as you possibly can anyway.

But, for larger organizations serving the enterprise or mid-market that haven't yet rolled out tools like Gainsight PX, Pendo, or any other app monitoring solution, a manual health score is worth the effort.

Your solution has health score metrics built into it. For example a CRM can be scored against new contacts created, new opportunities created, user adoption.

These are true outcome driving success metrics within your application. If you have the ability to login as your customer, your CSM can utilize a manual health scorecard and review the structure, adoption and utilization of the solution and score it.

Using this score, the CSM can develop initiatives to support the customer in driving the business outcomes forward. Additional training perhaps, assistance in a reconfiguration, or general business consultation where they share what other customers are doing.

How are others identifying cx health without automated tools?


------------------------------
Ryan Creamore - Infobip
Director of Customer Success, North America
------------------------------