What AI Can’t Tell You About Customer Decisions & Churn

All right. Good afternoon, everybody. Thank you for joining. My name is Natasha Narain.

I'm with Iceberg IQ and it's my pleasure to be joined by Ruben Rabago, my dear friend and former colleague for the session today. Something we both feel very strongly about. Ruben, do you mind sharing a little bit about yourself? Actually, I don't want to talk about myself.

That's a very easy thing to do. So instead, what I think I'd rather do is tell you a little bit story because you can find me on LinkedIn. You can go to customerrev.com. It's my personal brand.

I want to tell you a real story. So back in 2014, I joined this little company called Gainsight. It had about a million dollars of revenue, maybe 75 customers. And every once in a while, we'd get these reports.

They were incredible, thoughtful, thick reports, thorough reports more than anything. And they were these churn and win-loss analysis reports that were delivered by this company, this firm called Eigenworks. And it was led by this gentleman named Alan Armstrong. And at the time, our CEO, Nick Mehta, did what I thought was pretty bold.

He sent an email to the whole company, said, you know, I know these have some really sensitive information in it in terms of just the emotion. But everybody has to read these emails or these reports, these really thick reports. He said with one caveat, read it without blame. And that was really important.

That mattered, especially if one of those were your accounts where the customer was revealing all of their human emotion. And this was raw. This was customers that had gone through disappointment, that had had some unfortunate kind of experience. And you'd see your name on that report.

It wasn't a fun time to receive that. But what really caught my eye was this aha moment that I had at Gainsight, which was reading this report, it was here was a customer that had chosen us. Their expectation was to be successful. And somehow along the way, we lost them.

And I think that was the piece that no signal was telling us we were so advanced in all these health scores for our customers, but we couldn't even get to the root cause of why these customers were churning. And so when Ashwin and I had the opportunity to begin planning and writing the customer professional's handbook, customer success professionals handbook. Alan Armstrong was the very first person we asked to be a contributor to because we had learned so much about the customer journey and the customer experience that we asked Alan to write. And he was so excited to do so.

And literally just a few days before we were about to submit our final manuscript to Wiley, our publisher, Alan passed away unexpectedly. I miss him dearly. He's a beautiful man. And I am so honored and privileged that his words are still held and will live on in this book.

And if you all haven't read that piece, it's a gold mine. But I'm actually more excited because the folks at Iceberg IQ and my buddy here, Natasha, have continued the work of Eigenworks. Thanks, Reuben. So as you mentioned, we did work together at Gainsight.

I was also a recipient of those reports. And so many of our wins and losses and churns are Eigenworks would go out and interview them post decision and then give us all the feedback. And as a recipient, I actually had a situation where I was able to win back a deal as a result of one of these reports. It turns out that the customer or the prospect didn't want to come to St.

Louis for training, but they didn't realize they hadn't told me that. And they didn't realize that we now had an online offering. So I was able to go back and actually get that deal done, which helped to prove the ROI for the program itself. So here we are seven years later when I found out that Alan passed away in 2019, I was pretty devastated.

I reached out and offered my condolences and that led to conversations where we ended up spinning up Iceberg as a continuation of Eigenworks. And we're just thrilled to be sharing with you what we're finding with AI and insights today around human intervention. Thank you for sharing that. And thank you for having me today.

Well, let's get a shift to you because today is about you all. And with a show of hands, audience participation is really important here with the show of hands. How many of you are 100% confident you know the real reason why your customers turn? How many of you know with certainty why your customers turn?

We got one guy, one person out here. I want to talk to you afterwards. All right. What do you think about that one, Natasha?

Well, it's super interesting. So what we're finding a Gartner actually put out a report in 2025 more on the win-loss side. But the study for the guy that they put out was 65% of CMOs are now making investments in these types of programs where they're learning buyer decisions. And so it's only a matter of time that we see that spreading over to churn analysis.

But in our practice of 17 years doing interviews, half of our practice is doing churn analysis and half is half is win-loss. Well, that's sort of surprising, right? But not really. And what's interesting is we're going to hear it all day long, going to hear it all year long, all decade long that there's a huge expectation that AI is going to help uncover and reveal some of these unknowns.

And I'm here maybe controversially to say that while AI is going to be awesome, there are just some things it's not going to be able to pierce. Let me explain a little further. So customer success, we have absolutely no shortage of data, right? We have signals, we have support tickets, we have recordings from Gong or Zoom, and we have data entries in the CRM and the customer success platform solution.

There's Slack, AI on top of that. Now we have agentic AI helping us respond or react to all those signals. And what it can do is it can tell us, hey, you should this up usage is down. Sponsors sort of gone silent, disconnected, help us do all of that.

But what it's fundamentally not being able to do is interpret meaning. And that's where we stand. We're going to be all better at reacting to signals. But what AI is still not helping us do is understand the meaning.

Because humans are still required to uncover the why. I think that was a big thing that I learned almost 10 years ago with Eigenworks. And why did the sponsor go silent? Why did the customer begin to lose confidence?

Why did onboarding stall? Because rarely, the answer to those questions, it's just not going to come up in a single thread of data. And it can't be fully explained. And unfortunately, us as solution providers to our customers are sitting here, we're reading this book, we're going through this journey with our customers.

And sometimes we get to the end. And it's a surprising ending. Those green customers churn. Really looking well, good looking healthy customers end up churning.

And it's a sad story sometimes at the end. And what's disappointing, I think for all of us who've been in that situation, is that story was being written right in front of us as it was happening. And we missed it. And we missed it.

So what do leaders do? They I know I've been one, I am one. We like doing more. So more process, more dashboards, more tools, more AI, because that's going to fix it all right.

And effort goes up, but predictability sometimes does not follow. So what ends up happening at this point is if the data is not declaring what's happening, it still results in unknowns. Because AI is not going to be joining you in a room, and maybe you do have it, maybe it is recording you when you go to your customer on site, EBR, you're in their boardroom. It's not going to sense the tension in that room.

It's not going to be able to pick up when your executive sponsor is saying all the right words, but is thinking something completely differently. It's not going to pick up what isn't in the signals when the CFO behind the scenes is challenging the CCO. Why the spend? Why the budget?

What's happening here? This is where the data doesn't tell the story, and you have to dig in deeper, because ultimately, it's the outcome that we need. And at the end of the outcome, our people, our humans, today humans are still making the decision to buy, to implement, to use your solution. And even if your champion is a huge champion, he or she still has to convince their humans in their company to leverage the new workflows, the new tools, the new products.

There are still humans making the decision to renew, to expand, or to churn. And that's why I'm excited to be sitting here next to Natasha, because Iceberg IQ, from long ago, when I was a customer twice now, you all have figured out this hack. You've gotten into where AI can't. So why don't you tell us more?

Thanks, Reuben. So what Alan created was this methodology. It's really about the interviews, right? AI can help with the analysis, but we still feel like the interviews themselves are something that requires the human aspect.

He developed a methodology called Buyer's Hero, and it follows the framework of, I don't know if anybody knows, Joseph Campbell, his approach to the story arc being a hero's journey. So the buyer's really on this journey or this quest, and they change as a result of it with the decisions that they make. So when we're doing an interview, they're typically 30 minutes long, where we're interviewing one decision maker from the organization on behalf of our client. We're spending time upfront really profiling them.

What was their vision? What were they looking to achieve? What was the sales to service handoff like? What was the onboarding experience like?

What was that entire journey? What went well? Where did they struggle with the team, with the solutions? Where did things start to go wrong?

What was the last straw? And who did they ultimately go to and why? With wind loss, it's a little bit different because it's more about the perception of what they're expecting to achieve. So the methodology is the same, but the questions are different.

But this is what we look to uncover, and we have a number of interview techniques that we use to really get them into storytelling mode and open up. But this is really the framework around the churn interviews. We've done probably 3,500 interviews now, maybe more. We've worked with hundreds of companies.

So we're speaking to people and organizations all day long. And we are noticing that in the last couple of years, there's so much pressure around doing more with less. There's so much consolidation going on, not only with M&A, but also consolidation within the tech stack that organizations are using. Buyers are being asked on what they can use internally that it might not be as robust, but a solution.

And as Ruben said, every decision, whether it's renew, purchase, cancel, it's humans that are at the heart of those decisions. So you might be asking, why can't AI do this for you? And so here's a few examples of what we find when we're doing human interviews. One is it's the customer's voice that you're really capturing.

So we're allowing them to hold the space to reflect on their entire journey, and especially in the low touch or the tech touch segments, they might not have ever had any human interaction with you as the vendor. We're asking them what could have changed the outcome. And that's something that you're probably not going to get through gong calls or CRM reason codes, but we're looking to uncover that. And they're generally more candid with us as a neutral third party.

I had an interview just recently last week where I asked them why they canceled and they said because working with that vendor was a nightmare that I needed to wake up from. And then we went through the whole experience. So they can be pretty raw sometimes in these conversations. We're gathering competitive Intel, so it's not only about what what push them away from you or why they left, but it's also about what pulled them towards the other vendor.

So we see that decisions to churn are often made 12 to 18 months before the renewal date, but your renewal playbooks might be 30, 60, 90 days before the renewal date. And those competitors might be standing up a POC or something in parallel. They might be doing creative tactics around the commercials to help win them over. But those are the things that we uncover, which helps you in CS, but also helps the product marketers.

And these are validated by a third party expert. So what we've noticed lately is that executives are coming to us saying we need a better narrative or a stronger narrative for the board. We can present the churn reasons, but we're not actually telling the real story. And so we'll go in and do that digging deeper.

It sometimes resolves internal disputes, so it could be support that appears that it's the issue, but really it might be the underlying product. And as I mentioned, it's trusted by a third party. So here's an example of, I think, an ideal state of a happy path or a renewal path. And we've all seen these journey slides.

These are idealized for us. But what we do in our interviews is we're typically doing anywhere from 10, 15, 20 interviews in a dataset, in a study, or per quarter. We uncover the next path, which is the churn. So we'll identify patterns across the interviews about where these friction points lie and what you could be doing to help.

And so here's an example. Maybe they, well, actually, I'll give you a real example we had with a client where they said that the account managers and the CSMs were so focused on upselling, but they hadn't received core value in the initial offering or the initial implementation or purchase that they made. And so that's just a classic example of having that nuance or knowing when it's appropriate to be upselling or offering more instead of focusing on the established value. So our last slide, before we open up to questions, is just that because we've done so many of these, we've seen successful programs and not so successful programs.

An example of successful ones were when you have the executive stakeholders, and that's usually from the CEO down who's really invested in this program. Example is when we were at Gainsight with Nick. It was a top-down approach with a spirit of transparency. It wasn't about blame.

Everyone was there to learn, and it does really help. Defining a focus area for the research, so that could be a segment or a product or maybe a stage within the journey for each study. Interviewing decisions from the last quarter. So our clients would give us, say, all the churns that fit a particular profile for the previous quarter, and then we go out and book interviews, right?

And that's where we have human and tech blended to book the interviews, but then the interviews are human-led and then the analysis is a combination. But anything prior to the last quarter, the data just becomes stale. People might have moved on. They might not remember all the details.

The competitive landscapes change. So that's one of our best practices. And using the findings in your VOC strategies as well as your board meetings. We had one client where they had a section in every board meeting, which was Iceberg IQ findings, where they were sharing the win-loss and the churn findings.

Assigning actions as you receive every report. So it can be sometimes overwhelming when you're receiving all this data. So having a plan in place and someone owning it that they're identifying actions, putting that in a spreadsheet, but having an owner, an executive owner. So an example for one we recently did, which was an EHS company where they had executives assigned to every action and it was a pricing strategy that they needed to implement.

They had some AI enhancement requests. So having that as a rolling thing as opposed to at the end of the quarter going, "Okay, now what do we do with everything that we've learned in the last quarter?" So with that, we'll open it up to questions. We have a couple minutes left. There's one question on Slido.

It says, "What makes Iceberg IQ do such incredible job with customer interviews, win-loss interviews and reports?" Let me guess, is that Evan? No, I think it's just the experience we've had. And we really love the work. We're open to the hard work.

We're very curious by nature as well. And we just, we really enjoy all the conversations that we have. A couple of points. We do have a table behind just at the entrance to the Innovation Theater.

We are giving away copies of Ruben's book, if anyone would like one. We are also giving away Friends of Ours wrote the Customer Success Talent Playbook. I'm not sure if you've seen this. Author by Swati Garg and Julie Fox and many others.

So they've left a number of copies for people to pick up. And we're also giving away Eddie Cooler. Any final questions? Oh, we have one.

[inaudible] So the question was around the approach for booking the interviews and how do we make sure that we're successful. So we have a team dedicated to this work. It's almost like an SDR function, but actually a couple people in our team are actors, which is interesting because I think they're just used to rejection and they're very resilient. But it's a combination of phone and email touches.

The fourth or the fifth attempt is an executive touchpoint. We do offer a gift card or a charitable donation. We run that through Black Hawk Networks. So it's $100 is what we find is a threshold.

Unless it's a senior executive you're pursuing, then you'd go higher. And we do send a survey after every interview to ask why they took the call. And surprisingly, 36% of them said it wasn't about the incentive. It's actually that they just want to help the organization, even though they canceled.

You're welcome. Okay, we're at time. Great. Thanks, everyone.