Find the 80% of Customer Signal You’re Missing

Hello, Gainsight Pulse. I have to say this is my first Pulse. I'm so excited to be here. It's been really fun, hopefully first of many.

I want to quickly introduce myself. Thank you, Natasha, for the quick background, some of my travel stories. So I'm the founding account executive at QuadSci. So I was the first sales hire to our organization.

And when I joined QuadSci almost two years ago, since then I've been having conversations with customer success leaders every single day. And even at Pulse, I've had a ton of conversations across different areas of customer success. And I have to tell you all, I came to a really important realization about customer success, and I'm saying this as a salesperson. I realized, and I'll admit here on stage today, that your job is a lot harder than mine is.

Yeah. Came to that realization, wanted to share, it's harder than it has ever been to be in customer success. And we're seeing this is a problem across the industry, right? It's already difficult to be in customer success, no matter what year it is.

But in the year 2026, the data shows us that 65% of software companies are experiencing flat or declining NRR. So this is a problem across the entire industry. And the question really is why, right? It's 2026.

We have more data. We have agents. We have AI. But something really is missing from the way that we go out and support our customers.

So why is this happening and how do we actually fix it? So I want to talk to you all about the current state. What are the data points that your team actually uses and references on a day-to-day basis when you're talking to your customers? So we're thinking about things like your health signal, your NPS score, interactions that are logged into your CRM.

And this is really only about 20% of the total signal that you're getting from your customers. So where does the rest live, right? Is it coming from product usage? And that really is the key that I want to talk to you all about today.

So when you're operating with only a handful of data points that are also interpreted by a person, there's not much data for our AI to use and actually turn back to our customers to work with them in a way that's relevant, right? So the vast majority of the signal and what most customer success teams and sales teams are not using at all today is your product usage. Think about it like this. Your customers are talking to you every single day through the use of your product.

And they're actually giving you unbiased data for you to then use to go back and talk to them in a way that resonates, in a way that makes sense based on their experience with your brand. So when we unlock this signal, this 80%, we can finally have a full picture of our customer's view. But I have some good news and some bad news about product usage data specifically. The bad news, I'll start with, this is a massive data set.

It really is typically the largest data set that you have at your company. The other bad news is that it's really not meant for people to interpret easily, right? It also typically lives with product, with engineering. Most CS team sales teams are not thinking about product usage data.

We may be looking at really high level data points like logins, key feature interactions. But again, this is interpreted through a human saying this is what's important. So we want to take the guesswork out of that. So QuadSight, this is how we work with our customers across that 100% of the signal.

And we really focus on telemetry data. So when we can unlock everything, we see some pretty crazy results with our product. And I want to talk about one of our customers in particular that came to us with a serious churn problem, not unlike many software companies right now. And a lot of what they were doing was that 20%.

So they came to us and said, okay, what else can we do? We're curious about this product usage data, but we don't know how to work with it. And the great thing about product usage data is that it is the perfect use case for machine learning. Tons of auto-generated, unbiased data points that machine learning and AI can sift through to find patterns, to find key features that are associated with growth and with churn.

So this customer in particular, we worked with them over the course of about eight weeks. And we built our predictive models that looks for five different outcomes for our customers. That's high growth, medium growth, stable, contraction, and full churn. Because we know that the binary method of good, bad, growth, churn, it's really not enough in today's day and age.

We need more context. So we worked with them to get these key outcomes, and we actually found $5.2 million of surprise churn. And this is a company that now is at $185 million in ARR. They were about 135, 140 when we started working with them.

So it's a significant portion of their ARR that they did not know about. They didn't know it was at risk. And typically, whenever I talk about this customer story in particular, the question I get is, "Okay, what is surprise churn? How do you define surprise churn?" So I talked about what this customer was doing prior to working with QuadSci.

And they were looking at and identifying their customers based on green, yellow, and red. Sound familiar? So we looked at all of our QuadSci insights. What were the predicted risks across their customer base based on our models?

And we compared it with what they were doing. What did they identify as actually healthy customers? And that was the delta. So the fact that we're rooted in product usage data, everything we do is revenue-based, it's really easy to quantify the ROI of this.

So $5.2 million in surprise churn. And our predictive accuracy was 97% of surprise churn of churn identified one year in advance. So this is what happens when you get that full signal. You get this level of predictive accuracy.

And this is not unique for our customers. We consistently see well over 90% accuracy for a 12-month forecast window. So that means for your team, you're now telling them a year out, "Who are their risky customers?" And we all know we built up CS Playbooks and methods for saving our customers. But typically we deploy them or we find out about it just too late.

We don't have enough time. So QuadSci allows you to get that forecast window to actually impact what we're seeing. And this customer, they went from roughly $135, $140 million in revenue to $185. They saw 40% year-over-year bookings growth because they were able to identify the risky accounts, do something about it, stop the bleeding, so to speak, and actually shift their focus to growth.

So not only are we able to find risk and the patterns associated with it, but we can also find the hidden growth opportunities that your team does not know about today. And I mean, listen, this is a salesperson's dream, right? Tell me what customers are going to grow and tell me what they need. What's next for them.

And we're here with Gainsight. So of course I want to talk about for those Gainsight customers out there, how does this integrate with your Gainsight ecosystem, right? There's a lot of power across Gainsight. And where we focus our product usage data, this is-- there's a lot of product analytics tools out there.

Gainsight PX is a great example of one of those. So this is what's collecting all of that auto-generated customer data. And this is what we use for our analysis as the core. We marry that up with revenue data coming from places like your CRM, maybe some support tickets, all those other signals, that 20%, to then deliver it into Gainsight CS.

So we create this loop within Gainsight where all of those rich signals from product data are actually showing up to your customers or to your team in Gainsight CS where they live and work every single day. So I want to talk quickly about some of the new things for QuadSci and how we think about interacting with this data. Because it's not enough to just have the account assignment. Is it high growth?

Is it at risk? What are the signals? But it's actually, what do I do about it? And then how can I interact with this data to get deeper and deeper insights to this data set?

So conversational QChat. QChat is our generative AI capability that merges all of the rich quantitative data. That's coming from our AI models and brings in context from your CS playbooks, best practices for your customers, even external signals. All of this can be surfaced and interacted with through our conversational QChat functionality.

And then it wouldn't be Gainsight Pulse if we didn't talk about MCPs, right? This is the hot topic of the week. So QuadSci has our own MCP. This is all about making sure this data, all of this rich context, is showing up in front of your teams where they're working every single day.

So all of that product data can inform their experience in a cloud environment, in the Gainsight MCP. All of this works together seamlessly. So we are going to take a quick look at the platform. I have some videos that I'm going to share with you.

This is a quick snapshot of QuadSci. And if you want to see more, you can come find us at our booth. We'd love to talk to you after this. But we're going to look first at the highest level view that QuadSci gives you of your business.

So leaders are typically wondering what's happening across their territory, right? So on the QuadSci homepage, we're seeing the highest level revenue retention forecast. You get a 12-month rolling forecast with QuadSci. You can quickly identify where is the risk and where do all of our accounts fall?

How many are in growth class 5 churn risk? How many are in growth class 1 high growth? So we can see the entire business, the trends that have been happening over time. And then we can quickly look at what accounts are falling into each of these categories.

We can sort by renewal date, right? Growth class, we want to focus on 5s, 4s. Where do we want to focus? And then we can dive into the customer level.

So QuadSci goes all the way from across the business and then down to the individual account level for you to dive into the insights. So we can see the history of this customer. How have they evolved over time? What are we predicting for this customer 12 months out?

And what are the signals that are associated with this pattern? So what are the different features, interactions that are driving this prediction? We can also compare how they're interacting versus other customers like them. So maybe similar to their industry.

How has their adoption changed over time? And comparing across companies, across their own company, all of that. And if there's any operations folks in the house, this is typically where ops really likes to spend some time within QuadSci. So what you're looking at is the analysis across high growth and medium growth customers.

So when you're thinking about programmatic enablement for your teams, right? What are the features and patterns of behavior that we can drive across our customers that will likely put them in a better place? This is what you're seeing here. And this is based on your product, of course.

So we can see all of the different features that are associated with high growth. And again, enable our CSMs and our account managers to drive those behaviors in their customer base. And I'll give you an interesting example. One of our customers, they found a really interesting trend based on this analysis where they saw a specific feature usage at a specific frequency of a specific persona was one of their top growth signals.

So they created it as an organizational KPI for their entire team. They said, you need to get this persona in this part of the platform this many days a week. It will have this impact on your customers. Pretty cool.

So when we look at the individual customer level, and here we're going to focus a little bit more on QChat, that generative AI functionality that we talked about earlier. So at the customer level, like we said before, it's really not enough to just have the growth class, the signals that are driving it. We need their next best action, right? We have to close the loop.

What does my team actually do with this information? So on the right hand side, you can see the QChat recommended actions. This is all, again, rooted in product usage data, very specific to this customer and the trends that we've been seeing with them in particular. So the CSM can go in and say, okay, what is the next best action?

How do you recommend I actually do this? We can take it all the way to the end to actually get this out in front of your customers. And finally, we're going to go through the MCP. So a lot of you will find this interface familiar in something like Clod.

Well, we all know that data is critical for these LLMs, right? If they don't have accurate data or rich enough data, they can't really do much. They can't inform what happens next. So QuadSci is providing the deep product usage level context to inform the LLM and help you build out your workflows within Clod.

So again, this works across different MCPs. But today we talked about the problem across software. NRR is struggling. Everyone is dealing with this.

And I want to challenge you to think about your telemetry data. Think about product usage data as the superpower that you have yet to unlock. And QuadSci helps you do that. So it interacts with all of your Gainsight ecosystem.

And we want to make sure that all this ends up in front of your team where they will see it, where they will use it every single day. So that's it for me. If you want to come see us at our booth, you can find us on the Golden Path to the Puppies. You'll see us on your way there.

But we'd love to talk to you all more. Thank you so much for having me, Gainsight. And I hope to see you all soon. Thank you.