ArticleHealth ScoringOctober 8, 2019
How to Score Customer Health Image

How to Score Customer Health

By Dan Steinman
How to score Customer Health is a subject near and dear to my heart and also critical to the long-term success of any Customer Success team.  In fact, it’s really just an extrapolation of what you’ve all been doing for years by coding customers as red, yellow, or green.  As you all know, that model is better than nothing but it has severe limitations:
  • It’s almost always manually updated
  • The criteria tends to be very subjective
  • As you grow, there tends to be LOTS of unscored customers
  • Without digging, it’s hard to tell why a customer is whatever color they are
  • Most CFOs would probably not trust it to drive an accurate renewal forecast
Having said all that, the stoplight scoring metaphor does move us in the right direction, which is to apply a health score to every customer.  If you could do this consistently and automatically, the results would be extremely valuable and everyone knows it.  I’ll talk about the value proposition later but, for now, let’s assume that there’s a way to do this (and there is) and talk about what you need to do to get there in case you ever do own an application that can execute on the automation, analytics, and reporting you need.

Defining "Customer Health"

First off, you need to decide what “account health” actually means.  If a customer’s score is 48 (scale of 1-100) or RED or C (scale of A-F), what does that mean? Here are some questions to ask yourself as you go through this thought process:
  1. Is the score a predictor of renewal/churn? Should it be?
  2. How quickly can it fluctuate (RED one day, Green the next)?
  3. Does it take into account not only the value you are providing to them, but also the value they are providing back to you?
  4. What are the various elements that make up the overall score (this is key)?
  5. Should any of those elements be subjective or do they all need to be objective and concrete?
  6. Are all the elements equally important?
As you go through this process, you’ll definitely ask yourself many more questions like these.  The process is a little bit painful, but fun and a very healthy exercise even if you’re not going to automate it.  You will ponder many important aspects of your customers’ health and formulate some important thoughts on each of them. Beware of how many people you have involved, however, because this is all about opinions, not proofs.  And, as everyone knows, opinions are like noses - everyone’s got one.

Customer Health Inputs

Let me share a few of the elements of overall customer health that you should at least consider as you ponder this topic:
  • Overall usage of your product (obviously)
  • Usage of the stickiest features of your product (not as obvious)
  • Depth of usage - % of product used
  • Breadth of usage - # of licenses
  • Growth of the account (could be $$ or licenses)
  • Length of time as a customer (especially if you are trying to predict renewal)
  • Discount level (Huh? Yes, it could be important)
  • # of renewals completed
  • # of upsells completed
  • % of spend that is non-recurring (are they spending $$ to become a better customer through training and services?)
  • Survey results
  • How often do they call Support?
  • Marketing participation – references, speaking engagements, case studies
  • Product feedback
  • Community involvement
  • Invoice history
  • Length of time between end of onboarding and first renewal (for new customers)
  • Executive relationship
  • Overall relationship
  • Are C-level execs users of your solution?
That’s 20 elements right there and I know for sure that you could add 10 more quickly without a lot of thought.  The point is that there are so many things that affect overall customer health.  The picture is never quite complete.  Notice we never even touched the social media components here.  The challenge for you is to narrow your list down to a manageable number that would be helpful even if you had to do it manually for only your top 100 customers.  And then, when you do take the leap to automate it, you’ll have the option to expand the number of elements to include more automated ones and fewer subjective ones. As I’m sure you noticed, some of the options listed above are pretty important AND purely subjective (e.g. executive relationship).  The good news about the subjective ones, I’ve found, is that they don’t typically change very often.  A quarterly review will typically be adequate.

Multiple Customer Health Models?

One further complication that you’ll figure out very quickly is that you will probably want more than one scoring model.  It might make sense to have a different model for customers in the first 90 days than those who are more than a year old.  Or, you might want to score your high-touch, strategic customers differently than you do your no-touch, self-service customers.  These are refinements that you can do at a later date once you have the foundation built.  Just keep in mind that this thought will hit you at some point in this process. The value proposition here is enormous.  A healthy, growing recurring revenue business may very well end up in a place within 3-5 years where the recurring revenue (renewals) is larger than the new business sales.  As this happens, the requirement to do much more accurate renewal forecasts will go way up.  And keep in mind that the expectation of your CFO, and rightly so, will be that the renewal forecast should be significantly more accurate than the new business forecast. Good and timely account health scores will logically drive accuracy into your renewal forecast.  Also, a trusted account health score will help prioritize the activities of your Customer Success and/or Account Management teams.  These are expensive resources and prioritizing them and making them more effective has tremendous value to the enterprise.

The Evolution of Customer Health

I know that the picture I’ve painted here is of a long, arduous, and oft-changing journey.  That’s true but believe me when I say that it’s both a fruitful and fun journey.  As Customer Success organizations mature, this will not be an option so you may as well start early.  I think a good analogy is to the world of Marketing Automation. Account health scoring is akin to lead scoring.  And the MA folks will tell you that getting customers to do lead scoring seven years ago was like pulling teeth.  Today, many companies are buying their Marketing Automation tool specifically to do lead scoring and those that are serious about digital marketing are doing lead scoring from day one.  I guarantee you that the same will be true in the Customer Success world for account health scoring, and the grace period for getting there is very short and will get shorter by the month. So, jump in now and get the process started.  There may be no urgency now but there will be and you will be ready.  You may even find yourself leading the charge.  Trust me when I say that if your CFO or CEO gets wind of this capability and that it will greatly improve their ability to accurately forecast renewals, your time will get really short, really fast.  Why not be a step ahead of them for a change?
Picture of Dan Steinman
Dan Steinman GM, Gainsight EMEA GM, Gainsight EMEA
7 Comments

Leave a reply

Your email address will not be published. Required fields are marked *

  1. Tim Chingos
    Apr 1st, 2014Reply

    Dan – insightful and concise approach to scoring customers. I’d also add that folks shouldn’t be afraid to test their opinions/assumptions against actual data when available (i.e. what was the executive relationship like for accounts that churned?). Thanks for sharing with customer success community. Cheers!

    • Dan Steinman
      Apr 2nd, 2014Reply

      Great feedback, Tim. Thanks for sharing. You are absolutely right – the long-term goal HAS to be to score customer health and refine that process so it becomes highly predictive. And the way to do that is to start with some assumptions and then check those assumptions against actual results.

  2. […] How to Score Customer Health – Dan Steinman […]

  3. […] Well, we already have a Customer Success metric like this, and it’s called aCustomer Health Score. […]

  4. Arvind
    Mar 20th, 2015Reply

    Dan, thanks for a really well written post. I have a question regarding the relative weights of the Health Score factors. Do you have any advice on how to go about developing the weights? Is trial/error/judgement on historical data OK, where you calibrate to current customers and where you think they should be, or do you have a more data driven recommendation? Cheers!

  5. Barb
    Feb 28th, 2017Reply

    Good read. Very helpful. Appreciate you taking the time to write this up and give guidance on how to approach and think through the measurement. Thanks.

  6. Chris Sparshott
    Aug 20th, 2018Reply

    The thing about Customer Health Score is it is not a measure of Customer Health.

    It is a measure of the Vendors success at meeting defined touch points which they believe has some influence on recurring revenue. Don’t get me wrong, a health score will give you trending data about churn but it is not a customer health score. We should flip our brains and come at this from the customers perspective.

    1. Did the client implement the project that the solution was designed to deliver?
    2. Did the client implement the project where the vendors’ solution was one small part?
    3. Did the software pass the clients ROI gate to determine if the software should be used in future projects
    4. Did the software add value to the end users that the client was reaching out to – the reason the project was executed in the first place
    5. Do we have resellers involved and do they have the skills to help the customer.
    etc etc etc

    The above points focus on projects and supporting metrics. These have a far greater impact on the churn rate than the number of upsells. Especially true if you sell into large System integrators or large organizations performing a transformation project.

    Which leads to Customer Health Scores weightings need to changing by deal size and client size. The weightings to calculate the health is not the same for each.

    If I were to take it a stage further, I would look to add Myers Brigg personality score against the buyers and stakeholders as that also defines the way in which those individuals are satisfied. If your messages are designed to be delivered to extravert type personalities you miss the silent majority, like the CFOs and Engineering Leads. (generalization there..)

    Food for thought!

Shares