Software & Market Strategy
Customer Success Metrics for Software Companies
Lauren B. Jones
CEO & Founder, Leap Advisory Partners
Lauren B. Jones
A software company I advise lost their second-largest customer last quarter. ARR impact: $180,000. When I asked the customer success team if they saw it coming, they said no. When I looked at the data, the signals had been visible for six months: login frequency had dropped 40%, support ticket volume had tripled, their executive sponsor had left the company, and they had skipped the last two quarterly business reviews.
The data was there. Nobody was watching it.
This is the customer success measurement problem in B2B software. Teams track churn and celebrate retention, but both metrics are lagging indicators. By the time you see churn, the customer has already decided to leave. By the time you celebrate a renewal, the decision was made months ago. The metrics that actually matter are the ones that tell you what is about to happen, not what already happened.
Churn rate is completely useless for managing individual customer relationships because it measures an outcome that already occurred. Monthly or quarterly churn rate is the metric that gets reported to the board. It is important for understanding the business at a macro level. But it is completely useless for managing individual customer relationships, because it measures an outcome that already occurred.
The customer who churned this month started their exit process 3-6 months ago. Their frustration built over time. Their usage declined. Their engagement with your team decreased. By the time they gave notice, the relationship was already over. The cancellation was just the paperwork.
What you need is a set of leading indicators that flag at-risk customers while there is still time to intervene. These indicators are not exotic. They are based on observable behaviors that correlate with retention or churn. The challenge is building the system to track them consistently and act on them promptly.
A customer health score aggregates multiple signals into a single indicator of relationship health. The best health scores are predictive: a declining score today predicts churn 60-90 days from now, giving the CS team time to intervene.
1. Product usage. This is the strongest predictor of retention. Track daily or weekly active users as a percentage of licensed seats, feature adoption breadth, and usage trend over the last 90 days.
2. Support trends. High support volume can indicate deep engagement or deep frustration. Track ticket volume trend, severity distribution, time-to-resolution trend, and CSAT scores on resolved tickets. A shift from "how do I do X?" to "X is broken again" is a dangerous pattern.
3. Engagement cadence. Track QBR attendance, responsiveness to CS outreach, community participation, and executive sponsor engagement. A customer who starts skipping QBRs and takes a week to respond to emails is almost certainly evaluating alternatives.
4. Expansion signals. Customers who are expanding (adding users, adopting features) are healthy. Customers who are contracting are at risk. A customer who expanded in each of the last three years and is not discussing expansion this year has changed their perception of your value.
5. NPS trajectory. NPS matters less as an absolute number and more as a trend. A customer whose NPS dropped from 9 to 7 is at higher risk than one whose NPS rose from 5 to 7. The direction reveals the direction of the relationship.
Not all metrics are equally predictive. These have the strongest correlation to renewal outcomes based on patterns I have observed.
90-day usage trend is the single best predictor. A customer whose usage has declined 20% or more over the last 90 days is 4-5x more likely to churn than one with stable or growing usage. This metric should trigger an immediate outreach from the CS team.
Executive sponsor continuity. When the executive sponsor (the person who championed the purchase) leaves the company, churn risk increases dramatically. Executive sponsor departure preceded churn 60% of the time if no new relationship was established within 30 days.
Time since last meaningful interaction. Not a login. A meaningful interaction: a QBR, a strategy call, a feature request conversation. Customers who have gone 60+ days without meaningful engagement are at elevated risk.
Support satisfaction trend. A declining trend in post-ticket CSAT scores predicts dissatisfaction even when the customer is not complaining directly.
Your CS dashboard should answer three questions at a glance: Which customers are healthy? Which are at risk? What action should the CS team take today?
Data sources. Pull from product analytics, CRM, support system, and NPS platform. Connect and normalize so the health score updates automatically.
Visualization priorities. Lead with a heat map or tiered list: green (healthy), yellow (monitor), red (at risk). Sort by revenue impact within each tier.
Alert thresholds. Define triggers: usage dropped 20% in 30 days, executive sponsor left, no meaningful engagement in 45 days, CSAT dropped below threshold on last 3 tickets. Include recommended actions with each alert.
Metrics without action plans are dashboards without value. Build playbooks for each health score tier.
Green customers (healthy). Focus on expansion. Playbook: quarterly strategic reviews, proactive product recommendations, executive alignment, referral requests.
Yellow customers (monitor). Focus on stabilization. Playbook: increased touchpoints, usage deep-dive, targeted training on underutilized features, executive sponsor check-in.
Red customers (at risk). Focus on intervention. Playbook: immediate executive outreach, candid conversation about concerns, fast-track issue resolution, value realization review, retention offer if appropriate.
The most sophisticated software companies have repositioned customer success from a cost center to a growth engine. The same health scoring framework that predicts churn also identifies expansion opportunities.
A green customer with high usage, growing user count, and engaged executives is a prime expansion candidate. The CS team should identify these proactively and coordinate with sales.
The math supports this shift. Acquiring a new customer costs 5-7x more than expanding an existing one. The gross margin on expansion revenue is higher. And expanded customers have higher retention rates because they are more deeply embedded in your product.
Customer success metrics are not just about preventing bad outcomes. They are about identifying the best opportunities in your existing customer base and systematically capturing them.
The four strongest churn predictors are the 90-day usage trend (a 20%+ decline makes churn 4-5x more likely), executive sponsor departure (preceded churn 60% of the time when no new relationship was built within 30 days), time since last meaningful interaction (60+ days without a QBR, strategy call, or substantive conversation elevates risk), and declining support satisfaction scores on recent tickets.
Build a health score from five components: product usage (active users as percentage of licensed seats, feature adoption, usage trend), support trends (ticket volume, severity, CSAT), engagement cadence (QBR attendance, email responsiveness, executive sponsor involvement), expansion signals (user count and feature adoption trends), and NPS trajectory (direction of score matters more than absolute number). Aggregate into green/yellow/red tiers sorted by ARR.
Churn rate measures an outcome that already occurred. The customer who churned this month started their exit process 3-6 months ago. Their frustration built gradually, their usage declined, and their engagement decreased long before they gave notice. By the time churn appears in your metrics, the relationship was already over. Leading indicators like usage trends and engagement cadence flag risk while intervention is still possible.
Build tiered playbooks triggered by health score changes. Green customers get expansion-focused activities (strategic reviews, product recommendations, referral requests). Yellow customers get stabilization outreach (increased touchpoints, usage deep-dives, targeted training). Red customers get immediate intervention (executive outreach, candid concern discussions, fast-tracked issue resolution). Each alert should include a specific recommended action so CS managers can respond immediately.
Ready to build a health score for your customers? Download the Customer Health Score Template. It includes scoring criteria for all five components, threshold definitions, and playbook outlines for green, yellow, and red accounts.
Download the Customer Health Score Template
Lauren B. Jones is the CEO and founder of Leap Advisory Partners, with 28 years of experience in staffing technology. She helps staffing agencies, PE firms, and software companies build technology that actually works.
