meaningful stories

Predictive Modelling for B2B Churn: Building Your Retention Arsenal

Apoorv Nandan
CTO & Co-founder
November 9, 2023

Predicting churn is a complicated affair because by the time you zero in on specific churn signals (also known as lagging signals), it’s already too late and customers are way past the horizon. Leading metrics like social churn signals offer a superior alternative to pro-actively track and prevent churn by mapping social interactions and identifying preemptive indicators that convey a likelihood of churn.

A social signal can be anything from a specific event to an action, message, or intention that surfaces a likelihood of churn in the future. These indicators, although subtle, are easy to identify and act on if CSMs pay attention to client conversations. Because if a customer is about to churn, it will show up in their actions or messaging. No customer churns without a reason, there are enough signs if you know where to look.

What are Social Churn Signals?

Predicting churn will never be as straightforward as a customer telling you that they want to quit. However, even the most non-vocal clients drop subtle hints via social interactions that may imply an impending churn. These signals are called social churn signals.

To put it plainly, social churn signals are patterns in the social fabric of client interactions that unveils an inherent dissatisfaction or problem. It can be anything from a negative review, declining social engagement, or a click on the competitor’s website. Oftentimes social churn signals don’t necessarily come from inside the product but rather from the outside that are impossible to track using regular customer data analytics tools. They are easy to miss unless you’re aware of the likelihood of their occurrence.

However, there are tools like Amplemarket that can help track these signals by monitoring the social media presence of clients. These tools focus on your competitors and allow you to see if someone from your client’s team is interacting with your competitor’s product. This is what they call competitive intelligence, that enables business stakeholders to reach out to such customers with personal messaging and identify friction points within their ecosystem that compels their customers to look for alternatives.

Lagging vs Leading Churn Signals: Which ones to use?

All user KPIs correlated to customer churn can be grouped into two categories, leading and lagging indicators. While leading indicators are precursors to the action itself, lagging indicators represent the result of the action. Now, for people looking to study the cause and effect of actions on their business, both are equally important. However, lagging indicators rarely help if you’re looking to prevent churn, because by the time you read these indicators, it’s already too late and churn has either occurred or is inevitable.

One way to differentiate between leading and lagging indicators is that leading indicators are objective in nature and hard to measure since they are designed to surface a user’s feelings. Social signals are leading indicators as well and one of the best at predicting churn. Here are some of the other leading indicators you can use along with social signals to significantly lower your churn.

  1. NPS: NPS or Net Promoter Score is essentially a score assigned to users based on their answer to the question “how likely are you to recommend our product to friends, family, or someone with the same problem?” Clients that answer with an 8 or above are advocates and anything less than that is a defector.

  1. Product Activation: Product activation can be a leading or a lagging indicator based on your goals. In the context of preventing churn, product activation is a leading indicator as it can help shorten the time it takes for a customer to activate their licence and start effectively using the platform.

  1. Social Signals: Social signals are simple interactions or peripheral events that indicate a likelihood of churn. They are usually subtle and barely hostile. However, they offer very early signals that a customer might churn.

  1. CSAT: CSAT or customer satisfaction score is a comprehensive score assigned to users to measure how happy they are with the product. You can couple CSAT scores with lagging indicators like retention and growth to uncover patterns and trends.

  1. Usage: Product usage is a phenomenal indicator for user retention and churn prevention. If customers rely on your product more, they are less likely to churn unless there’s a big change in their organisation (like new leadership or a merger, which again is a social indicator).

Now if you’re looking to measure the effect of your efforts made to reduce churn, you might want to steer your attention towards lagging indicators like Renewal rate, growth rate, and churn rate. But since our goal today is to understand how one can use social signals (a leading indicator) to identify churn signals and prevent them, we will steer clear of lagging indicators.

Although we do realise that quantifying your churn prevention efforts is equally important, it goes beyond the scope of this blog. You can read more about them here.

Social Signals Preceding Churn

Social signals are less intuitive because unlike other churn indicators they usually don’t surface when things are visibly going south and are therefore harder to correlate to churn. Here are some less talked about social churn signals that can predict the likelihood of churn with high accuracy.

Interaction with a Competitor

Any interaction that your client does with your competitor is a sign that there’s friction somewhere in your customer flow, or your client has an issue that they can’t get rid of even after repeated customer support interactions. For instance, consider a marketing agency that has been using your AI-based direct response copywriting tool for over a year now. You’ve never had any friction with this client. But you recently found out that some of the performance marketers from their team are trying out your competitor’s free plan.


Leaving this issue unattended will most probably end up with the client leaving you for your competitor. These interactions are traceable through website cookies, and social media activity of your client and can be addressed by personal outreach messages.

Change in Organisational Structure

Any organisational change, no matter how small, is worth paying attention to. Although most things remain the same when someone new takes the helm, some things are bound to change. For instance, your POC might change and may have disagreements with regards to the inherent value of the product. Or a new manager might want to switch to tools they’re more familiar with. Or the teams using your product might take this opportunity to finally adopt new workflows and tools.

Regardless of the scenario, it is better to read these signals and stay on your toes for any contingencies that might arrive. Establishing a transparent communication channel and not letting the communication break is the key to avoiding friction in such scenarios.

A Feature Request From a Quiet Customer

When was the last time you made a feature request on a whim?

Nobody does it for fun or experimentation, when a user makes a feature request there’s usually a motivation behind it. PLG vendors are more likely to face such instances and while it’s true that it’s not viable to grant every feature request, some of them are truly genuine and are worth considering.

When a user that’s usually quiet takes the trouble to finally reach out to you for a specific feature, it’s a huge sign that there’s an issue that needs immediate attention. And when you deny that request without further investigating the underlying motivation, it shows that you’re inconsiderate.

Investigate feature requests thoroughly before making a decision to avoid such instances.

Odd Customer Behaviour

While a feature request from a quiet customer is an anomaly, it’s more of an oddity when a client suddenly goes silent. And by silent we do not mean, not asking questions, or zero customer support interactions. It’s when they stop responding to your messages, when you start paying attention towards the issue.

However, it would be a bad idea to jump to conclusions based on this one signal. If a POC stops responding to your texts, it may be because they’re on a vacation, or they have been moved to a different project, or they have shifted to a new organisation. Whatever it is, you should consider reaching out to someone else on the team that uses your product. If you don’t fetch a response and there’s a drop of usage across the team, then it’s probably time to employ retention strategies.

Personal Differences with a Customer

Synergistic conversations indicate a healthy client-vendor relationship. No other user touchpoint has a deeper impact on client relationship than the interaction between the CSM and the client POC. You have to understand that there are personalities behind the product usage and any event that causes tension between a manager and a client may come in the way of the next renewal.

A strong customer relationship is absolutely non-negotiable if you want to retain your customers and such instances should be treated with caution. The best course of action would be to assign a new CSM to the client for healthier and frictionless interactions.

Value Perception Drops

Value perception isn’t the actual value your product is capable of delivering to the client, it’s what the client feels your product is delivering to them. It depends on factors like feature activation, onboarding flow, and UI/UX.

Your client might see a sudden or gradual decrease in the perceived value of your product due to changing priorities and goals. And if they see this trend continuing long enough, they will cancel their subscription. You can gauge these signals by looking at several other indicators like declining usage, users unsubscribing from the product newsletter that they’ve been actively engaging with for over a year, or not engaging with your organic content.

Therefore, it’s better to stay in the loop and discuss upcoming goals and priorities of your client during meetings.

Change in Account Information

B2B customers have standard SOPs for regular operations and it’s not everyday that you see updates in their account details. Change in contact information and payment details is a major redflag and may indicate a search for alternatives.

Such signals should be treated with caution and reaching out to these clients may uncover the underlying reasons for their search.

How To Use These Signals to Prevent Churn?

You took the advice and you started paying attention to your clients and their social interactions, and while at it, you noticed some churn signals.

Now, what next?

Well, you need to act on them and take consistent action to make sure that whatever is bothering them is taken care of. You can use some of the following strategies to ensure higher retention.

  1. Don’t Just Interact, Communicate

The best way to prevent churn using social signals is to never let them appear in the first place. Don’t wait for your client to reach out to you before you start interacting, anticipate potential friction points and communicate consistently, even if you think that the customer is satisfied.

Ask them if there’s something you can do to improve the utility of the product for them, tell them about any major updates that you’re going to roll out, and make sure that they feel acknowledged.

  1. Use Personalised Messaging

This might sound cliche, but personalised messaging is a powerful retention tool because of two reasons-

  1. It shows that you understand client’s problem or are at least trying to do so
  2. It also makes them feel valued

Found a long-time user interacting with a competitor? Reach out to them with a personalised email and ask them whether there’s something they wished for in the product that's missing.

A client just went silent on you? Try contacting someone else from the team and check user analytics platforms to see if there’s a drop in usage across the team or is it just one person.

Not every social churn signal requires personalised messaging, but reaching out to your clients whenever you sense friction isn’t a bad idea.

  1. Collect and Analyse (Social) Data

Social events are often neglected when it comes to business intelligence. The ethos that is prevalent in B2B SaaS circles is that social events have little to no influence on customer behaviour. However, business intelligence and customer behaviour aren’t mutually exclusive.

People are greatly influenced by their friends, friends of friends, competitors, and colleagues. Vendors can unearth vast amounts of untapped insights by tracking customer activity across social channels like websites and social media. Churn intent may arise at any stage of the customer journey and may go unnoticed for quite a while if one ignores social churn signals.

One way to ensure that you’re not missing out on hidden social signals is to use emotional intelligence (an AI subset that unearths emotional insights from remote or live interactions). Platforms like Lightbulb and Cogito extract emotional insights from regular client interactions and offer a more holistic view of the user’s pain points when used alongside social churn signals.

Harnessing user data, especially when there are multiple silos to draw data from, is intimidating, and yet the only alternative to losing revenue is to churn.

That’s what Crunch solves for you.

Join the waitlist now to know more.

Tell meaningful stories

Read similar blogs

UPGRADE YOUR ANALYTICS EXPERIENCE

The ultimate intelligent data layer for every business

Get Started With Crunch