There is a window in most B2B renewals where intervention is still cheap. The customer has not made up their mind. Their team has not started evaluating alternatives. The friction is still friction, not a decision. What we have seen across our early-access work is that this window tends to open four to seven weeks before the renewal conversation, and the signals that mark its beginning are sitting in your support queue and email threads right now.
This piece is about what that window looks like and why it is so easy to miss until it closes.
What the window actually contains
When we talk about a six-week window, we are not describing a moment when customers suddenly announce frustration. It is subtler than that. What appears in this period is a shift in the pattern of how customers communicate about your product. The topics change, the framing changes, and sometimes the volume of communication drops in ways that are easy to read as healthy engagement when they are actually disengagement.
The support queue signal is usually the most concrete. You see a drift from questions that presuppose continued use ("how do I set up X for our Q2 workflow") toward questions that suggest review ("what is included in our current plan" or "how do we export all of our data"). The email signal is typically tonal: threads that used to carry collaborative problem-solving language shift toward requests for documentation, for escalation paths, for clarity on contract terms. Neither of these individually looks alarming. As a pattern across four to six weeks, they tell a different story.
Why this particular window matters
Six weeks is not arbitrary. It is roughly the lead time needed for a CSM to do anything meaningful before a renewal conversation. If you find out about friction two weeks before the renewal date, you are mostly in damage-control mode: you can acknowledge the problem, offer a credit, promise a roadmap item. What you cannot do is actually change the customer's experience of the product before the conversation that matters.
At six weeks, you still have time to open a real conversation. You can bring in someone from the product team. You can offer a workflow review that addresses the friction they have been signaling in tickets. You can proactively surface the value that your product has delivered in areas they have not thought about recently. None of that is available to you at two weeks.
The catch is that six weeks before the renewal is exactly when most CS teams are focused on the accounts that are obviously at risk, the ones with open escalations, the ones that have already said something is wrong. The accounts that are quietly drifting show up as green in most health dashboards right up until they do not.
The anatomy of a quiet signal
The signals in the six-week window are quiet by nature. They do not look like trouble if you look at them one at a time. A single ticket asking about data export is normal. A CSM receiving a more formal email tone from a champion who used to write casually is easy to explain away. It is the combination, over time, that carries the information.
Some specific patterns worth watching. First, topic specificity in tickets: customers who are engaged with a product as a long-term part of their workflow tend to ask specific, operational questions. As they move toward churn, questions become more generic and more oriented toward understanding the scope of what they have versus what they need. Second, escalation aftermath: when a support issue gets escalated and then resolved, healthy accounts resume normal ticket volume. Accounts in the pre-churn window often go quiet after an escalation in a way that does not reflect resolution so much as withdrawal. Third, thread participation: the number of people a customer copies on email threads drops. Early in a relationship, threads are wide because the customer is investing in getting your product embedded in their team. Narrowing threads can reflect narrowing commitment.
None of these are definitive on their own. All of them together, emerging over a four-to-six-week span, represent a coherent behavioral shift.
Why CSMs miss it
The structural reason CSMs miss this window is simple: no one has time to reread six weeks of tickets for every account in their book of business before each renewal. A CSM managing fifty accounts cannot synthesize patterns across support, email, and call notes for all of them simultaneously. The information is there. The bandwidth to process it is not.
The second reason is that the signals are genuinely ambiguous at the individual level. A customer who asks about data export might be doing routine IT housekeeping. A narrower email thread might just mean the champion has a smaller team this quarter. Individual signals require context to interpret, and that context accumulates over time in ways that are hard to hold in memory.
There is also a cognitive pattern at work: CSMs tend to anchor their assessment of an account on the last direct conversation they had. If that conversation was positive, the account feels healthy regardless of what has been happening in the background. The signals in the six-week window are typically happening in channels that do not directly surface to the CSM: tier-one support tickets, email threads that do not include the CSM, call notes from the account manager. The CSM's mental model is based on incomplete information, and it stays optimistic longer than it should.
What you can do with six weeks
The most useful thing a CS team can do with this window is not to immediately schedule a call. The first step is understanding what the signals are actually pointing at. Is the drift in tickets about a specific feature gap? Is the tonal shift in email correlated with a particular point in their fiscal calendar? Is the narrowing thread participation happening in one team within the account or across the whole relationship?
Once you understand the shape of the friction, you can respond to it specifically rather than generically. A generic "checking in" call six weeks before renewal is unlikely to surface what the signals have been pointing at. A call that opens with "we noticed some questions coming through about your data export workflows and wanted to understand what you are planning for Q2" lands differently. It demonstrates that the CSM is paying attention to the account beyond scheduled touchpoints, and it opens a door the customer may have assumed was closed.
Not all accounts in the six-week window are going to churn. We want to be direct about that. Some of what looks like a pre-churn pattern is ordinary operational communication that clusters before annual planning periods. The goal is not to panic about every shift in ticket topic or email tone. The goal is to not miss the ones that are real, and to have enough lead time to respond when they are.
A note on what this requires
Finding the six-week window reliably requires a different kind of attention than most CS tooling is designed to provide. Health scores built around product usage give you a view of what customers are doing in the product, but they do not tell you what customers are saying about the product. Those are different data streams, and the one that tends to move first before a hard renewal is the communication stream.
This is early days for Sturdy. We are still learning which signal combinations are most predictive, which account types exhibit the clearest patterns, and how to present this information in a way that is useful to a CSM who is already managing a complex book of business. What we are confident about is that the window is real, that it starts in the support queue and the email thread rather than the health dashboard, and that six weeks of lead time is enough to do something meaningful with it.
