Tracking Product Adoption: A Critical Step to Customer Success
Although I don’t believe that product adoption is the complete story of customer health, there’s no denying that it is a critical aspect. It is the primary factor in the success of your customers and the key driver of most Customer Success activities. It certainly deserves lots of attention and an intelligent discussion of what to track and how.
Let’s start by talking about what to track. The easy answer is “everything” but that is not the no-brainer it sounds like. Unless you are fully prepared to sift through massive amounts of data in search for some gold nuggets or you already have your big data processes and algorithms all figured out, you may be better off with less data rather than more. Don’t get me wrong – if you have the ability to track everything, by all means do so. That historical data will someday be extremely valuable when you do start looking at it through a data science lens. So, by all means, track everything you can, but think long and hard about what to pay attention to in the near-term. Less is often more as you get started in this endeavor.
A simple way to approach this challenge is simply to sit down as a Customer Success team and talk about what the drivers of churn or at-risk customers are. I guarantee you that some of your instincts are correct. We’ve all dealt with those customers at renewal time who “aren’t getting enough value” out of your system. And we typically know, or they tell us, what they mean by that. For example, if your system requires users to login to use it, that’s an obvious place to start. Some period of time without a login should definitely trigger a red flag. Beyond that, you should analyze what it is you want, or expect, your customers to be doing with your product.
Let’s take an example of an application everyone understands – file sharing and collaboration. Think about the first few things you’d want your customers to do with that kind of application – uploading and downloading files. Let’s take marketing automation as another example. The basics here would include uploading, segmenting, and filtering lists, and sending out emails.
These kinds of activities are clearly the place to start. But, for most applications, these basics, while important, will leave you exposed to churn if that’s the extent of your customer’s usage. Your product is priced based on the value it can deliver through the deeper, more sophisticated features, not just the simple ones.
The next logical step then, is to think about your stickiest features. This is usually the functionality of your product that drives the biggest ROI. Think about which parts of your application, if used reasonably well, make it much harder to replace. As those of you in the Customer Success discipline already know, these are often harder features to get your customers to adopt, but they are where your true value lies.
Customers using these more complex, more valuable features are far less likely to churn, and far more likely to buy additional products from you when they come along. Back to our file sharing application example, these features might be things like file sharing and collaboration- features that start to involve other people and stretch your value beyond yourself. In our marketing automation application, this would be functionality like lead scoring and more complex nurturing campaigns, as opposed to one-step email blasts.
As we begin to think about how to track these customer activities, let me take a brief aside here to give you great hope for the results of this effort. Not only can the collection and analysis of customer usage data help you start to become proactive in your intervention with customers who look to be at-risk, but all of this data can also be put to great use immediately through the application of data science. For most companies, this is not something you can do yourself, but the expertise and the algorithms do already exist to use your usage data to predict both churn and positive behavior.
In simple terms, think of it this way – shouldn’t all of the data about all of your churned customers give you great insight into what the predictors of churn are? The answer is obviously yes. And not just in theory, but in practice. Similarly, all of the data about all of your renewals can tell you which elements are correlated to both renewal and upsell. This is very exciting because, as you well know, most CEOs are somewhat interested in what happened in the past and why, but are MUCH more interested in knowing what’s going to happen down the road six months.
Tracking usage data is typically done in one of two ways:
- Instrumentation of your product from within your product. This is usually done such that usage activity is tracked by user, by company, and by time, and this data is then fed to a database or data warehouse.
- Real-time collection from your web application. This usually involves putting some tracking codes on your web pages to track page views and/or specific activities on a page (like pushing the “submit” button for example).
Both methods are valuable and valid and can even be complementary. Both also pose some challenges you should consider. If your product has already been instrumented, that’s good news. Most SaaS applications have some level of instrumentation built in from the very beginning. If this is true, the good news is that there’s history already available to you. This is really important because, from the lens of Customer Success, it’s the trends in the usage data that are really important. What happened today, or this week, are vaguely interesting by themselves. But what happened this week, compared to the last 30 weeks, becomes extremely valuable.
The challenge here is often that the initial instrumentation has not been improved upon. Your newer products may have no instrumentation at all unless your Product team has been really diligent in these efforts. The other challenge is that this data usually needs to be moved somewhere (an application designed to analyze it for example) in order to be really valuable. Moving and transforming data is not a new concept so there’s no rocket science involved in doing this, but it doesn’t happen by magic. Someone who knows what they are doing and has the right tools, will almost certainly need to be involved.
By the way, moving this data into your CRM system, while a step forward, will not really help you solve this problem because CRM systems are not usually designed to do any time-based analysis.
The challenges with the second method - real-time collection of usage data - are twofold:
- There’s no history. After one day of collection, you have one day of history. After one week, you have one week of history. Better than nothing but it will take some time before it becomes really valuable.
My bottom line message to you is this – in order to successfully manage your customers in a recurring revenue business, you HAVE to know what they are doing with your product. Find out if it’s being tracked and stored somewhere. If so, start working to get your hands on it (and asking for more to be tracked). If not, then ask your Product team to start requiring instrumentation of key user activities and find someone who can immediately start helping you with real-time tracking. Google Analytics, Mixpanel, and segment.io are a few of the many ways to get started here.
The reward here is nothing less than survival if you are counting on renewals or recurring revenue so, in this case, “he who hesitates is lost” has never been truer.
Get to it. Now!