Renewal Prediction and Forecasting Best Practices for CS Teams

Sometimes, your customer tells you they have a problem without actually saying it. I came across something in my Customer Success class that made me look at customer data differently. Before this, I would probably look at numbers and think: “Okay… these are just metrics.” But I'm beginning to understand that, for a CSM, data can tell a story. Imagine you have a customer who has been actively using your product. They're logging in regularly. They're engaging with the product. Everything seems fine. Then suddenly… Their usage starts dropping. Maybe their logins are down. Their engagement has reduced. They're no longer using a feature they previously relied on. The customer hasn't complained. They haven't sent an angry email. They haven't said, “We're struggling.” But the data is already whispering: “Something has changed.” And that's where a proactive CSM pays attention. Instead of waiting until the customer completely disengages, you look at the numbers, notice the change and reach out. Not to ask: “Why aren't you using our product?” But to ask: “Is everything okay? Is there anything happening on your end that we can help with?” Because behind that drop in usage could be a new challenge, a change in their team, difficulty using a feature, or simply a customer who isn't getting the value they expected. And this is one thing I'm really beginning to appreciate about Customer Success: Data helps you know WHAT is happening. Empathy helps you understand WHY. You need both. I'm learning that being a good CSM isn't just about looking at numbers. It's about knowing when those numbers are telling you that a customer may need you. Still learning. Still connecting the dots. And definitely starting to look at customer data differently. 📊 For those already in Customer Success, what customer metric do you pay the closest attention to—and why? #CustomerSuccess #CustomerSuccessManagement #CSM #CustomerExperience #DataDriven #LearningJourney #CareerGrowth

Most account health scores are more complicated than they need to be, and less accurate because of it. I talked about this with Sara Wyman, Founder and CEO of Stackpack, on Across the Funnel Podcast. Her team runs vendor and spend intelligence for finance and IT teams, and by month two they need a clean answer on which customers are fine and which ones need a call. They keep the account health scoring down to two signals: → Data under management: what percentage of a customer's actual contracts are sitting inside the product → Usage: logins, procurement workflows run in the tool instead of a spreadsheet, people asking the embedded AI assistant a question because they wanted to, not because someone told them to Neither signal is about sentiment. Both are about proximity, how much of a customer's real work has actually moved inside your product, and how often they come back once it has. The habit doesn't survive without somewhere to look. A CSM juggling CRM notes in one tab, product usage in another, and a spend export somebody emailed last month isn't going to catch the contract-file percentage sliding for six weeks straight. Somebody needs both numbers sitting next to each other, updating on their own, with something that flags the moment either one moves. Most health scores measure how a customer feels. This measures where they actually live.

✨ A single customer story can be powerful—but patterns across many customers can be transformative. The Law of Large Numbers explains something every Customer Success team eventually learns: one customer can mislead you. A thousand customers usually won’t. In statistics, the Law of Large Numbers tells us that as the number of observations increases, the average of those observations tends to get closer to the true expected value. Customer Success is surprisingly similar. 📊 Imagine you have 10 customers: 2 are extremely happy. 😊 5 are moderately satisfied. 🙂 2 are struggling. 😟 1 is threatening to churn. 🚨 Looking at those accounts individually can make it difficult to understand what's really happening. But increase the sample size to 1,000 customers, and the picture becomes much clearer. 🔍 You might discover: → Customers using Feature A have higher retention. 📈 → Customers who don't complete onboarding within 14 days are more likely to churn. ⏳ → Accounts with consistent engagement generate more expansion revenue. 💰 → Certain industries consistently struggle with the same part of the product. 🏭 This is where Customer Success becomes more than relationship management. It's about understanding the difference between an anecdote and a pattern. One customer saying: "This feature doesn't work for us." is valuable feedback. 💬 But if hundreds of customers report similar issues, and the data shows that those accounts also have lower adoption or retention, you now have something much stronger: A signal worth investigating. 🚦 And this is why modern Customer Success teams need to think statistically. Not every unhappy customer represents a product problem. Not every churn is a failure of the CSM. The larger the customer base and the better the data, the easier it becomes to separate signal from noise. 🎯 The real opportunity is to stop asking only: "How is this customer doing?" and start asking: "What are our customers collectively telling us?" Because sometimes, the most important customer insight isn't hiding inside one account. It's hiding inside the average. 📊 #CustomerSuccess #DataDriven #CustomerExperience #SaaS #Analytics #DataScience

Your customer health score is probably not what your CFO wants to see. They want to know: “How much revenue did Customer Success protect or create?” That means replacing: ❌ Adoption increased 18% ❌ NPS improved ❌ Engagement went up ❌ Customers attended more QBRs With: ✅ $2.4M in churn risk identified ✅ $800K in expansion influenced ✅ $1.1M in renewals protected ✅ $350K in revenue at risk recovered AI makes this easier. You can connect product usage, support activity, sentiment, renewal data, expansion signals, and account history to identify financial patterns at scale. The future of CS reporting isn't: “Here’s what our customers are doing.” It’s: “Here’s what those behaviors mean financially.” Question: Should Customer Success own a revenue number? #CustomerSuccess #AI #CFO #SaaS #CustomerExperience #Revenue

A VP of product I talked to last week built a customer health dashboard. Part of assessing each customer's health was to look at the positive patterns that led them to renew. Side note, I love to see Product leaders acting like GMs / CEOs of their products. In my experience, far too few product leaders take accountability for the *business* side of their products. I digress... Great instinct on the part of this product leader, and there is one tweak I'd make to their analysis. Their approach didn't account for survivorship biasl. During WWII, the US military was losing far too many aircraft in battle. To solve this, technicians began mapping damage on aircraft that returned from battle. They planned to reinforce the aircraft with armor in the areas where they were most often damaged. What their plan missed was an analysis of the planes that didn't return from battle. The casualties. The ones that were damaged in areas that prevented them and their pilots from returning home. A statistician named Abraham Wald convinced his colleagues to take the exact opposite approach. They needed to reinforce the areas that WEREN'T damaged in battle. Often we look at customers who make it to year two, three, and beyond, and try to replicate their successes, but we look past the failure modes of the ones we lost. It's harder to track down this information and it isn't all cleanly measurable (a clean dashboard gives us all the warm-fuzzies, right?). But what I've found across two and a half decades of customer-facing work is that there are factors at play that we won't fully understand until we dig deeper. Factors that explain the failure modes in a much more granular and nuanced way. In WWII we couldn't go back and ask the pilots who were shot down where they were hit, but we could infer it based on the planes that returned from battle. In tech we can make those inferences, too. What valuable areas of the product were our customers NOT using? And behind the data, qualitative discovery (i.e., asking the customer what happened and why) fills in the rest of the story. How do you analyze the success of your product? Are you falling victim to survivorship bias?

For years, I believed customer retention was more art than science—something driven by intuition rather than numbers. But that changed when I dove into historical user data with the goal of finding the 'Aha!' moment—the point when a customer decides to stay. Using logistic regression, I discovered a striking pattern: customers who completed 3 integrations AND added 5 team members in their first week had an extraordinary retention rate, with churn shrinking to just 2%. This was more than a number; it was a revelation. It showed that the secret to loyalty lies in very specific early actions, a moment that could be engineered. Armed with this insight, teams could shift strategy from guesswork to precision—building onboarding processes that guide users through these pivotal steps. For me, it was a lesson in how data uncovers hidden stories—stories that transform how we engage and retain customers forever. Have you experienced a defining moment in your customer journey that changed everything? #CustomerInsight #DataScience #RetentionStrategy #SaaS

Customer Success doesn’t have a data problem. It has a decision problem. Most CS teams already have access to plenty of information: 📊 Product usage 💬 Support activity ❤️ Health scores 👥 Stakeholder changes 📝 Customer feedback 🔄 Renewal signals But having more data doesn’t automatically create better outcomes. The real challenge is turning fragmented signals into clear decisions and meaningful action. A stronger approach is: Identify what matters → Understand the context → Take action → Measure the impact Because the goal isn’t to collect more customer data. It’s to make better decisions with the data you already have. What’s the biggest challenge in your CS organization today: getting the right data or acting on it? #CustomerSuccess #CustomerIntelligence #CSLeadership #CustomerData #CustomerExperience #CustomerHealth #SaaS #CustomerRetention #CustomerValue #SciqusAMS

I’ve been thinking a lot about multi-threading into accounts as a CSM. 🧵We know we need to have power users, champions, and an executive sponsor fully engaged if we want to protect revenue and hopefully expand as well. This is part of the role as an Enterprise CSM when your book size is a ‘manageable’ 10-15 accounts (or less). When your book size grows beyond 15 accounts or so I’m curious how you operationalize multi-threading? ❓ I’ve seen various experimentations over the years but not sure I’ve seen anyone confidently show they have this depth and breadth fully engaged before as book sizes grow. AI is certainly changing the game here when it comes to customer engagement so I’d love to hear what’s working for you that allows for human-to-human, personal connections across these personas at scale? Bonus points every time you type out segmentation ;) 💡

We call it the “customer journey”, but we measure it like it belongs to us, not the prospect. We have more ways to measure the customer experience than at any point in history: Impressions. Clicks. Open rates. Website behavior. Content downloads. MQLs. SQLs. Pipeline. Conversion. Revenue. And once prospects become customers, we keep measuring: Adoption. Utilization. Satisfaction. Expansion. Renewal. Churn. We add increasingly sophisticated attribution, analytics, and AI, and now we’ve built an extraordinary measurement apparatus around the customer. But look carefully at what most of it tells us: Whether we’re winning. Did they notice us? Engage? Buy? Use more? Renew? Expand? All legitimate question, but notice how they’re remarkably different from another question: “Is the customer winning in all of this?” Because customers aren’t trying to become MQLs, enter our pipeline, increase our utilization, or improve our NRR. They’re trying to accomplish something that needs to become easier. Faster. Less expensive. Less risky. More effective. So why don’t we apply the same measurement rigor to their progress toward value that we apply to their progress toward us? There’s an important difference between measuring the value we extract from a customer and the value a customer realizes from us. Here’s my point: maybe the customer journey was never really about how successfully customers move through our business. Maybe it’s about how successfully we help them move through theirs.

Your churn data is lying to you. Most Customer Success teams are still looking at "last login" dates and "seat utilization" as their primary health indicators. But in the era of AI, these are lagging metrics. By the time usage drops, the decision to leave has likely already been made in a boardroom you weren't invited to. The new era of CS isn't about monitoring activity; it’s about predicting intent. We are moving from a "Reactive Firefighting" model to an "AI-Driven Offensive" strategy. Here is how the top 1% of CS organizations are reshaping their playbooks: 1. **Sentiment Analysis at Scale:** AI doesn't just track if a customer opened a ticket; it analyzes the *tone* of every Slack message, email, and call transcript. It identifies "Executive Dissatisfaction" or "Competitive Mention" long before a CSM senses a vibe shift. 2. **The "Digital-Led" Global Scale:** You can’t hire your way to global expansion anymore. AI-powered agents are now handling the "how-to" queries and technical onboarding, freeing up human CSMs to focus on strategic business reviews (SBRs) and mapping software value to the customer’s Board-level KPIs. 3. **Predictive NRR:** Net Revenue Retention is no longer just a finance metric—it’s a data science challenge. AI models can now correlate specific feature adoption patterns with a 90% likelihood of expansion. If a customer hits "Feature X" and "Feature Y" within 30 days, the AI triggers an expansion playbook for the CSM automatically. The goal isn't to replace the human element of Customer Success. It’s to remove the "grunt work" so CSMs can actually be *Success Managers* rather than *Support Reinforcements.* In an AI-driven economy, your product is a commodity, but your ability to proactively deliver ROI is your moat. **The takeaway:** If your team is still spending 60% of their day manually updating CRM notes and chasing "ghost" customers, you aren't scaling—you’re stagnating. Is your CS team currently using AI to predict churn, or are you still waiting for the "cancel" email to arrive? #CustomerSuccess #AI #SaaS #CustomerRetention #NRR #GrowthStrategy #ArtificialIntelligence #CustomerExperience