Becoming a Data-Driven Customer Success Team

Our first question addressed how to define a data-driven Customer Success Practice. One of the main points of contention among our experts: "Does data replace the classic human decision-making tool, intuition?"

Chad and Bob thought metrics and insight should replace intuition and "hunches," while Lincoln argued that "data-driven" was a baseline from which to make critical decisions, but that data doesn't replace the decision or decision-maker (the human element).

Tom mentioned that data was helpful for making decisions about customer health, tactics, programs, and resource allocations – using data to inform decision-making along with using intuition to figure out which metrics to look at in the first place. All experts agreed that CSMs need to find ways to leverage quantifiable data to help them prevent churn, even if they still rely on "non-traditional" data sources like "gut feelings".

We asked our experts to weigh whether data or relationships matter more in the world of Customer Success. Todd thought that relationships were critical, above all. Nichole agreed, and pointed out that in her experience, NPS scores for relationships were more important than those for product. Chad disagreed, suggesting data trumped all else. Bob felt both were important, and Tom seconded that, also saying relationships were especially critical at the high end.

Nichole recommended a high touch approach for retention, including frequent customer visits, and our experts discussed what this would mean at scale. The verdict: relationships are most important at higher price points, and data is the foundation or "context" for strong relationships.

The experts weighed in on the practical aspects of communicating the value of a SaaS offering to customers. Bob suggested that using self-reports (where customers estimate the value being received) can be helpful. Other ideas: showing customers their results through objective metrics like usage, while acknowledging that high usage does not always mean high achievement, and that low usage does not equate to low value.

Tom offered that finding out exactly how your customers get value and then measuring that is crucial. You need to track what success means to your customer, not just what you think it should mean. One way to do this, Tom pointed out, is relatively straightforward: Ask them! Chad noted that customer advocacy is another good measure of whether your customer is getting value.

So which is better context – how a customer compares to their own past performance, or how they compare to other customers? Lincoln thought a customer's own track record is the way to go. Bob and Tom felt that both comparisons were important, but recommended focusing on a customer's own track record first and graduating to benchmarks later. Todd, Chad, and Nichole all felt that customers needed to define their own value and focus on achieving that, rather than look at historical or external benchmarks.

Usage is just one way to think about Customer Success, so we asked our experts to tell us which other kinds of data matter most. Loyalty (retention, advocacy, upsells) and NPS topped the list. Also discussed were: time to first value, other lifecycle and contract milestones, MRR, support data, billing data, content interactions, and meetings.

Lincoln, Chad, and Todd all pointed out that the "gut feel" inputs into customer health from CSMs and others who interact with the customers also have an important place in the equation.

This question explored the "so what next?" of Customer Success initiatives, looking at how the analytics can and should be actionable. Bob suggested that customer experience improvements could be a direct result of data analysis. Todd, Lincoln, Tom, Nichole, and other chat participants all talked about an early churn warning system. And productivity, feature development, upsells and new sales, and second-level onboarding were also mentioned.

The theme: quality data analysis can enable the "proactive save" by highlighting patterns of trouble that could be churn indicators, and can essentially super-charge your CSMs.

For SaaS companies just starting down the road of Customer Success data analysis, what are the most important considerations? Start small, choose a specific problem to solve, know your data, and know your biases were some of the valuable tips offered.

Bob, who lives and breathes data, pointed out the importance of rooting out biases. And Tom stressed how crucial it is not to silo efforts to prevent misunderstandings.

Can your customers benefit from the learnings you are gathering from their data and from your analysis of your customer data as a whole? What is the right role for transparency? Ideas from the experts included: monthly scorecards, quarterly business reviews, insight emails, and in-app messaging. The right channel and cadence seemed to depend on the customer's need, with the goal being to have the communication occurring as often as possible to make renewal a "non-event".

Lincoln Murphy pointed out that transparency is crucial – if you're scared to share your data findings with your customers, that's a problem.

Just how important a tool can predictive analytics be for customer churn? We know the likes of Verizon and Netflix are all over this practice, but how does it apply in B2B SaaS, if at all? Our experts didn't find consensus on this topic, with Todd raising a warning flag, Lincoln pointing out the importance of taking the right actions on findings, and Bob explaining that you need to carefully measure what you are trying to predict.

Should SaaS companies let a thousand flowers bloom or focus in on just a few key SaaS metrics? And if so, which ones? Our experts were all over the map on this, mentioning metrics that touch on sales, usage, sentiment, willingness to recommend, and overall engagement.

Join us for the next #CustomerSuccessChat on Tuesday, 10/21/2014 at 9a PT/12p ET. The topic will be High-Touch at High Volume: Automated Outreach for Customer Success!