Customer usage data can seem intimidating. As Customer Success professionals, we might not be used to working with it. Thankfully, it’s not as tricky as it seems – and it can be the key to nailing your customer journey!
I have a confession to make:
I used to hate data. I thought it was boring.
I’m interested in people, in psychology. Pivoting tables? Not so much.
But I’ve come around, and now I can admit it:
Not only is data useful – it’s actually pretty cool.
Once I understood how to use it better, it unlocked so many insights into my customers and what success looked like for them – so much so that it changed the customer journey map significantly.
The Customer Journey
Just so we’re all on the same page here, I want to get clear on what I mean by a customer journey map.
The customer journey (in the CS world) is the path a customer takes with your product. It includes all the interactions and touchpoints they have with your company, from post-sale through onboarding, adoption, expansion, renewal, and advocacy.
Customer journeys are important because they can help you understand your customers better and figure out what success looks like from their perspective. They can also help you build playbooks that enable consistency and allow you to scale over time.
How can customer usage data help me understand my customer journey?
Anyone who is using a CS tool (Like Gainsight, Churn Zero, Totango, SmartKarrot, ClientSuccess, etc.) is using data.
Whether it’s triggering a renewal playbook based on a customer’s renewal date, automating a welcome email to a new client, or alerting a CSM to a change in customer health, data is what’s driving that functionality. When it works, a CS tool dramatically increases how many customers a CSM can serve and allows them to be more strategic in their interactions.
But something that’s often neglected is usage data. Yes, it’s often factored into customer health, but usually only in the sense that more logins = good. Less logins = bad.
Those are important! But they don’t tell the whole story.
Usage data is a great way to see what success looks like using your product because it shows exactly what people are doing with it.
How often are they using the product?
What features are they using?
How long does it take them to complete a set of actions? What actions do people perform in what order to achieve success with a different use case?
What are they doing more of over time vs. what falls by the wayside after training?
Which customer usage data should I use?
To answer this question, it’s important to understand that there is no one-size-fits-all approach. What you use will depend on what your customer journey looks like, and what steps your customers need to take to achieve their goals. Some key usage data points include:
- Login frequency
- Duration of sessions
- Feature usage
- Order of actions performed
- Time in product
Reverse engineer success
While it can be fun to look for current trends, I find the best place to start is with historical data – if you have it.
Basically, this comes down to:
Look for what has worked in the past. Get people to do more of that.
Start with your “Successful” customers. Those who have renewed and have good customer health and/or a high NPS.
Look back at their usage data during the first 30-60-90 days with the product. Then take a snapshot of their usage every 90 days.
What did their adoption look like?
How often did they use the product?
What features did they use?
What does their continued successful usage look like?
Look for trends:
What did the successful customers have in common? Was it frequency? How quickly they were able to complete an action? Specific features?
Once you see the habits that led to successful adoption, you can build that into your onboarding and adoption playbooks. Customers are successfully onboarded when they ________.
Don’t keep it a secret from your customers. Tell them! “Our clients who are most successful at achieving X do ________. And here’s why…”
If you show them what drives success and why it works, they are more likely to follow suit.
Know what churn looks like
Now do the same thing with your churned customers.
Look at their 30-60-90 day adoption and usage metrics. How do those differ from successful customers? Where do they start to diverge?
Now you can turn this into a risk playbook. When a customer starts exhibiting churn behavior – isn’t using the product as much as would be expected, not using the features needed to achieve their goals… you now have the data to trigger an alert to your CSM that the customer is starting to go off track.
It is MUCH easier to save the customer at this point, than 3 months prior to renewal.
Help! My data makes no sense
If you have tried the above and the results are all over the place – where you can’t see any consistency in customer usage data and renewal/churn, don’t panic!
This is normal.
You just need to segment.
But this tells you something important! If all your customers don’t succeed in the same way, onboarding them the same way probably isn’t the most effective solution.
So start slicing and dicing.
Try looking at the same scenarios but by segment.
What do I mean by segment?
You can also look at it as cohorts. It’s just a group of customers who have something in common. The ones I use most often are:
- Use case
At one company I worked at, we had many verticals but two big ones were education and tech sales. For the education accounts, using the product 8 times a year – and never in the summer – made them completely happy and they renewed consistently. But the tech sales customers only found success if they were using the product a minimum of 3x per week – and often every day.
When looked at all together – especially with the other verticals on top, it made no sense. But as soon as we segmented by vertical, it came into focus.
If that isn’t happening after you’ve sliced and diced it a number of ways:
Data integrity refers to the quality of data that ensures it’s accurate and consistent. It also describes the process of ensuring that your data is clean, up-to-date, and in line with your business goals.
Data integrity also means that when you make changes to one part of your organization or product, those changes are reflected across every channel and platform in your organization.
(Hey! Stop laughing…)
Ever have another part of the organization forget to fill you in on something that affected the customer? Well, any change to the back end of the product that affects a feature might change the way it’s shown in your analytics. Only you won’t necessarily know that unless you know the context.
This is one of MANY reasons it’s helpful to leave the analysis to a professional – a data analyst, a BI team, a CS Ops professional… otherwise, you’re likely to misinterpret or get thrown by minor inconsistencies.
So what if my data isn’t clean?
It happens. It’s honestly the norm, at least in startups.
Here’s what to do:
Talk to your customers
Revolutionary, I know. But we all get so focused on data, scaling, and automation, that it can be easy to forget that we can ASK our customers what has worked for them. It’s what you do to build your first customer success journey, and it still works.
Work on cleaning it up
Unfortunately, you can’t do this alone. As soon as you can, start communicating the value of the clean data to your executive team. Make it clear that it’s essential to scale, and if it’s not clean, it takes time to clean up and/or untangle. So start yesterday.
Start from where you are
Once you have systems and processes in place to ensure you have clean data, start monitoring. You’ll still know more in three months than you do now. It’s better to wait and work with clean data than to use data that drives you in the wrong direction.
Put processes and safeguards in place to keep your data clean moving forward – and have it validated every 3-6 months.
Don’t rely on data alone
Customer usage data is awesome – but it’s one piece of the puzzle.
It can be misinterpreted. Even with the best of intentions, our findings often contain implicit bias.
We are looking at things through our own lens, and we are missing the things we don’t think to look for.
Beyond that, people are unpredictable and irrational, (except you of course) so understanding why they make certain decisions requires both qualitative and quantitative information.
And thank goodness for that. It means we still need humans to do this job.
Want to become a confident world-class CS leader? Let’s chat.